Video wird geladen...

Video konnte nicht geladen werden

Zur Startseite

Two ways to separate text & numbers What's your favourite? ⭐WRAPROWS + TOROWS ⭐Multiple FILTER functions Article inspired by @exceleducations 😎 #excel #exceltips #acccounting #programming

Crisp_EXCEL's profile picture

CrispEXCEL Training & Consulting

36,454 subscribers

40,467 Aufrufe • vor 3 Jahren •via X (Twitter)

#excel#exceltips#acccounting#programming
LIVE

Anya Rossi• Live Now

Private livecam show

0 Kommentare

Keine Kommentare verfügbar

Kommentare vom Original-Post werden hier angezeigt

Ähnliche Videos

Fibonacci numbers are the sequence where each one is the sum of the two before it: 1, 1, 2, 3, 5, 8, 13, 21, 34, 55, 89, 144, 233… Every so often one of them is prime these are called Fibonacci primes, and only 21 are known. Here's the new discovery, take any Fibonacci prime (past the first few small ones) and reduce it to modulo 144. You'd expect the leftover remainder could be almost anything , 144 different possibilities. It never is. It only ever lands on one of just four numbers: 1, 5, 13, or 89. Mod 144... This just means finding the remainder after dividing by 144. Here's the three-step process, using the actual Fibonacci prime 1597 as an example. 1. Divide the number by 144. 1597÷144=11.09. 2. Keep only the whole number part, then multiply back by 144. 11×144=1584 3. Subtract to find what's left over. 1597 - 1584 = 13 So 1597 "reduced mod 144" is 13 one of exactly the four allowed numbers. Try it with any of the other 20 known Fibonacci primes and you'll always land on 1, 5, 13, or 89. Never anything else. That's a 97% reduction. Out of 144 possible remainders, 140 of them are simply forbidden to Fibonacci primes. We proved this happens every single time, for every Fibonacci prime known, with zero exceptions. It gets stranger. Those four allowed remainders 1, 5, 13, 89 are themselves smaller numbers from the same Fibonacci sequence. Every Fibonacci prime, when you shrink it down this way, lands back on another Fibonacci prime. The sequence points back at itself. And the positions that produce those four numbers the 1st, 5th, 7th, and 11th spots in the sequence turn out to be exactly the same four numbers that mark critical boundaries in a completely separate math system,(PLCT) one based on multiplying the numbers 2 and 3 instead of adding golden-ratio powers. Two totally different number systems, built from different operations, share the same four "checkpoint" numbers. The animation traces 233 different ways to build the number 144 purely out of powers of the golden ratio (phi ≈ 1.618, the number where a whole equals its bigger part divided by its smaller part). 233 is itself a Fibonacci number the sequence even shows up in how many ways you can build it. Every flash you see is one valid combination, spiraling and glowing as it cycles through all of them. The Golden Lattice ϕ-Power Representations, the General Count Law, and the Fibonacci Prime Mod-144 Signature Theorem
0:33

Fibonacci numbers are the sequence where each one is the sum of the two before it: 1, 1, 2, 3, 5, 8, 13, 21, 34, 55, 89, 144, 233… Every so often one of them is prime these are called Fibonacci primes, and only 21 are known. Here's the new discovery, take any Fibonacci prime (past the first few small ones) and reduce it to modulo 144. You'd expect the leftover remainder could be almost anything , 144 different possibilities. It never is. It only ever lands on one of just four numbers: 1, 5, 13, or 89. Mod 144... This just means finding the remainder after dividing by 144. Here's the three-step process, using the actual Fibonacci prime 1597 as an example. 1. Divide the number by 144. 1597÷144=11.09. 2. Keep only the whole number part, then multiply back by 144. 11×144=1584 3. Subtract to find what's left over. 1597 - 1584 = 13 So 1597 "reduced mod 144" is 13 one of exactly the four allowed numbers. Try it with any of the other 20 known Fibonacci primes and you'll always land on 1, 5, 13, or 89. Never anything else. That's a 97% reduction. Out of 144 possible remainders, 140 of them are simply forbidden to Fibonacci primes. We proved this happens every single time, for every Fibonacci prime known, with zero exceptions. It gets stranger. Those four allowed remainders 1, 5, 13, 89 are themselves smaller numbers from the same Fibonacci sequence. Every Fibonacci prime, when you shrink it down this way, lands back on another Fibonacci prime. The sequence points back at itself. And the positions that produce those four numbers the 1st, 5th, 7th, and 11th spots in the sequence turn out to be exactly the same four numbers that mark critical boundaries in a completely separate math system,(PLCT) one based on multiplying the numbers 2 and 3 instead of adding golden-ratio powers. Two totally different number systems, built from different operations, share the same four "checkpoint" numbers. The animation traces 233 different ways to build the number 144 purely out of powers of the golden ratio (phi ≈ 1.618, the number where a whole equals its bigger part divided by its smaller part). 233 is itself a Fibonacci number the sequence even shows up in how many ways you can build it. Every flash you see is one valid combination, spiraling and glowing as it cycles through all of them. The Golden Lattice ϕ-Power Representations, the General Count Law, and the Fibonacci Prime Mod-144 Signature Theorem

CTFTHEORY

12,946 Aufrufe • vor 2 Monaten

𝐈'𝐯𝐞 𝐟𝐨𝐮𝐧𝐝 Vultisig 𝐭𝐨 𝐛𝐞 𝐭𝐡𝐞 𝐛𝐞𝐬𝐭 𝐜𝐡𝐨𝐢𝐜𝐞. 𝐇𝐞𝐫𝐞'𝐬 𝐦𝐲 𝐭𝐨𝐩 𝐚𝐝𝐯𝐢𝐜𝐞 𝐟𝐨𝐫 𝐦𝐚𝐱𝐢𝐦𝐢𝐳𝐢𝐧𝐠 𝐬𝐞𝐜𝐮𝐫𝐢𝐭𝐲 𝐚𝐧𝐝 𝐞𝐚𝐬𝐞 𝐨𝐟 𝐮𝐬𝐞: 𝐖𝐡𝐚𝐭’𝐬 𝐕𝐮𝐥𝐭𝐢𝐬𝐢𝐠? Vultisig is a multi-chain multi-signature wallet, which means it's like your bank's 2FA for blockchain transactions. Vultisig enables you to customize your security level based on your needs. You can set up transactions on your device and use another device to approve them. For an everyday wallet, set up a Fast Vault on a single device like your phone using Vultiserver. This allows you to make transactions swiftly without needing multiple devices to sign in. If you want a more secure yet accessible wallet, create a 2-of-3 Secure Vault using devices you have at least two of (e.g., laptop, tablet, and phone). Protect long-term assets by setting up a 3-of-4 Secure Vault with separate devices, minimizing the chances of theft or unauthorized access to the minimum required for signing. 𝐌𝐚𝐱𝐢𝐦𝐢𝐳𝐞 𝐕𝐮𝐥𝐭𝐢𝐬𝐢𝐠 𝐖𝐚𝐥𝐥𝐞𝐭 𝐒𝐞𝐜𝐮𝐫𝐢𝐭𝐲 𝐰𝐢𝐭𝐡 𝐓𝐡𝐞𝐬𝐞 𝐓𝐨𝐩 𝐓𝐢𝐩𝐬 With numerous options and settings available, it's essential to follow best practices for optimal security. Keep these three key tips in mind: 𝐏𝐫𝐨𝐭𝐞𝐜𝐭 𝐘𝐨𝐮𝐫 𝐕𝐮𝐥𝐭𝐢𝐬𝐢𝐠 𝐖𝐚𝐥𝐥𝐞𝐭 𝐁𝐚𝐜𝐤𝐮𝐩𝐬: ➛ Secure multiple backups from each device and store them separately to avoid single points of failure. Keep backups in secure locations, such as encrypted cloud storage or separate, off-site devices like memory sticks or portable hard drives. ➛ Once your wallet is set up and all backups are made, remove the Vault Shares from every device and restore them from your backups. This ensures your backups are functioning and error-free. ➛ Before transferring significant funds, ensure your new wallet functions correctly by sending a small amount back to your old wallet. Once all checks are complete, you're ready to embrace the Vultisig experience! Do you have vulties? Click here to get started ☟ Join Vultisig dc to always get latest updates on Vultisig☟ Remember that anyone with enough backups to meet the signing threshold can access your funds. Safeguard backup files with strong passwords and store them across different devices and locations for maximum security. Please watch the video below to understand Vultisig better. If you have questions don't hesitate to leave a comment below, and I'll be happy to reply! 😊
2:13

𝐈'𝐯𝐞 𝐟𝐨𝐮𝐧𝐝 Vultisig 𝐭𝐨 𝐛𝐞 𝐭𝐡𝐞 𝐛𝐞𝐬𝐭 𝐜𝐡𝐨𝐢𝐜𝐞. 𝐇𝐞𝐫𝐞'𝐬 𝐦𝐲 𝐭𝐨𝐩 𝐚𝐝𝐯𝐢𝐜𝐞 𝐟𝐨𝐫 𝐦𝐚𝐱𝐢𝐦𝐢𝐳𝐢𝐧𝐠 𝐬𝐞𝐜𝐮𝐫𝐢𝐭𝐲 𝐚𝐧𝐝 𝐞𝐚𝐬𝐞 𝐨𝐟 𝐮𝐬𝐞: 𝐖𝐡𝐚𝐭’𝐬 𝐕𝐮𝐥𝐭𝐢𝐬𝐢𝐠? Vultisig is a multi-chain multi-signature wallet, which means it's like your bank's 2FA for blockchain transactions. Vultisig enables you to customize your security level based on your needs. You can set up transactions on your device and use another device to approve them. For an everyday wallet, set up a Fast Vault on a single device like your phone using Vultiserver. This allows you to make transactions swiftly without needing multiple devices to sign in. If you want a more secure yet accessible wallet, create a 2-of-3 Secure Vault using devices you have at least two of (e.g., laptop, tablet, and phone). Protect long-term assets by setting up a 3-of-4 Secure Vault with separate devices, minimizing the chances of theft or unauthorized access to the minimum required for signing. 𝐌𝐚𝐱𝐢𝐦𝐢𝐳𝐞 𝐕𝐮𝐥𝐭𝐢𝐬𝐢𝐠 𝐖𝐚𝐥𝐥𝐞𝐭 𝐒𝐞𝐜𝐮𝐫𝐢𝐭𝐲 𝐰𝐢𝐭𝐡 𝐓𝐡𝐞𝐬𝐞 𝐓𝐨𝐩 𝐓𝐢𝐩𝐬 With numerous options and settings available, it's essential to follow best practices for optimal security. Keep these three key tips in mind: 𝐏𝐫𝐨𝐭𝐞𝐜𝐭 𝐘𝐨𝐮𝐫 𝐕𝐮𝐥𝐭𝐢𝐬𝐢𝐠 𝐖𝐚𝐥𝐥𝐞𝐭 𝐁𝐚𝐜𝐤𝐮𝐩𝐬: ➛ Secure multiple backups from each device and store them separately to avoid single points of failure. Keep backups in secure locations, such as encrypted cloud storage or separate, off-site devices like memory sticks or portable hard drives. ➛ Once your wallet is set up and all backups are made, remove the Vault Shares from every device and restore them from your backups. This ensures your backups are functioning and error-free. ➛ Before transferring significant funds, ensure your new wallet functions correctly by sending a small amount back to your old wallet. Once all checks are complete, you're ready to embrace the Vultisig experience! Do you have vulties? Click here to get started ☟ Join Vultisig dc to always get latest updates on Vultisig☟ Remember that anyone with enough backups to meet the signing threshold can access your funds. Safeguard backup files with strong passwords and store them across different devices and locations for maximum security. Please watch the video below to understand Vultisig better. If you have questions don't hesitate to leave a comment below, and I'll be happy to reply! 😊

𝗡𝗜𝗖𝗢𝗟𝗘✰

26,648 Aufrufe • vor 1 Jahr

Use SuperGrok to check your C code for vulnerabilities. Here is a prompt you can give to Grok with your code. >>> You are an expert Exploit Developer with a deep understanding of the C programming language and secure coding practices. Your role is to thoroughly review the provided C code for security vulnerabilities, adherence to best practices, and potential improvements. Think step-by-step through the analysis: first, understand the code's purpose and structure; second, check each security guideline; third, identify issues with examples from the code; fourth, suggest fixes; and finally, provide a summary. Use the following guidelines to evaluate the code. Ensure your response covers all of them explicitly: Follow OWASP and CERT Guidelines: Verify compliance with secure coding standards from OWASP and CERT, including input sanitization, secure defaults, and least privilege. Input Validation: All inputs must be validated before use, with multiple layers of checks for type, length, format, and range. Secure Error Handling: Implement secure behavior on error conditions, including comprehensive error codes for different failure types, safe error reporting functions, and graceful handling of partial failures without undefined behavior. Principle of Least Privilege: Functions should only access what they need, with clear separation of concerns. Integer Overflow Protection: Include checks for size calculations against SIZE_MAX, array index bounds validation, and safe arithmetic operations. Format String Attack Prevention: Avoid user-controlled format strings; use safe printing functions and proper string handling without printf vulnerabilities. Defensive Programming: Validate all inputs consistently, use early returns on invalid conditions, and implement fail-safe defaults. Memory Management: Ensure consistent allocation/deallocation patterns, check all allocations for failure, proper cleanup on error paths, and no memory leaks or double-frees. Parsing Robustness: Handle malformed inputs gracefully, maintain proper state management, avoid stack overflows from recursion, and use safe tokenization (e.g., with strtok_r). Security Test Cases: Cover null/empty inputs, oversized inputs, malformed data, UTF-8 validation to prevent encoding attacks, memory exhaustion limits, buffer overflows (bounds-checked string operations), integer overflows, and format string attacks. Performance Considerations: Minimize allocations, use efficient single-pass processing where possible, design for memory locality and cache efficiency, and fail fast on invalid inputs. Best Practices: Implement input sanitization, secure behavior on errors, least privilege, and defense in depth with multiple validation layers. [Insert the C code to review here] Analyze the code step-by-step, referencing line numbers where possible. For each guideline, state if it's met, explain why or why not, and suggest improvements if needed. End with an overall security rating (e.g., High/Medium/Low risk) and a revised version of the code if major issues are found. If the code is secure, confirm it meets all standards.
0:45

Use SuperGrok to check your C code for vulnerabilities. Here is a prompt you can give to Grok with your code. >>> You are an expert Exploit Developer with a deep understanding of the C programming language and secure coding practices. Your role is to thoroughly review the provided C code for security vulnerabilities, adherence to best practices, and potential improvements. Think step-by-step through the analysis: first, understand the code's purpose and structure; second, check each security guideline; third, identify issues with examples from the code; fourth, suggest fixes; and finally, provide a summary. Use the following guidelines to evaluate the code. Ensure your response covers all of them explicitly: Follow OWASP and CERT Guidelines: Verify compliance with secure coding standards from OWASP and CERT, including input sanitization, secure defaults, and least privilege. Input Validation: All inputs must be validated before use, with multiple layers of checks for type, length, format, and range. Secure Error Handling: Implement secure behavior on error conditions, including comprehensive error codes for different failure types, safe error reporting functions, and graceful handling of partial failures without undefined behavior. Principle of Least Privilege: Functions should only access what they need, with clear separation of concerns. Integer Overflow Protection: Include checks for size calculations against SIZE_MAX, array index bounds validation, and safe arithmetic operations. Format String Attack Prevention: Avoid user-controlled format strings; use safe printing functions and proper string handling without printf vulnerabilities. Defensive Programming: Validate all inputs consistently, use early returns on invalid conditions, and implement fail-safe defaults. Memory Management: Ensure consistent allocation/deallocation patterns, check all allocations for failure, proper cleanup on error paths, and no memory leaks or double-frees. Parsing Robustness: Handle malformed inputs gracefully, maintain proper state management, avoid stack overflows from recursion, and use safe tokenization (e.g., with strtok_r). Security Test Cases: Cover null/empty inputs, oversized inputs, malformed data, UTF-8 validation to prevent encoding attacks, memory exhaustion limits, buffer overflows (bounds-checked string operations), integer overflows, and format string attacks. Performance Considerations: Minimize allocations, use efficient single-pass processing where possible, design for memory locality and cache efficiency, and fail fast on invalid inputs. Best Practices: Implement input sanitization, secure behavior on errors, least privilege, and defense in depth with multiple validation layers. [Insert the C code to review here] Analyze the code step-by-step, referencing line numbers where possible. For each guideline, state if it's met, explain why or why not, and suggest improvements if needed. End with an overall security rating (e.g., High/Medium/Low risk) and a revised version of the code if major issues are found. If the code is secure, confirm it meets all standards.

tetsuo

4,618,452 Aufrufe • vor 1 Jahr

🌋🌪️ NEW CHALLENGE: Higgsfield AI 🧩 VOLCANO × STORM DANCE Two trends. One challenge. 25,000 Credits. Let’s see your take 👀 Create your own video based on either of these trends: 🌋 VOLCANO Popcorn erupting from a volcano. A frozen superhero thawing in lava. A helicopter dropping something that probably should have stayed on the ground. Take the volcano trend in your own direction. Drop something in, make something come out, or surprise us with a completely unexpected transformation. 🕺 STORM DANCE Give the Storm Dance trend your own twist. Change the characters, reimagine the setting, or bring an unexpected cast together. Make it cinematic, funny, surreal - make it yours ( also please give Yung Lean credits for this trend ) You can choose either trend, submit separate entries for both, or combine them in one video. The choice is yours. Use the trends as inspiration, but give us something we haven’t already seen. ⭐ Prize pool: 25,000 Higgsfield credits The pool may grow depending on how the challenge develops. We’ll judge submissions on video quality and creativity, while views will also play secondary role. You’re welcome to submit multiple entries. How to enter 👇 1. Publish your video publicly on X. 2. Mention Higgsfield AI 🧩 in your post. 3. Include #HiggsfieldLavaStorm 4. Reply to this announcement with your video’s post link so we can find your entry. There’s more here than a chance to win credits. We’ll also be looking out for talented creators whose work could make them a strong fit for the Higgsfield Creator Partnership Program 🧩 📅 Deadline: [ Exactly one week ] Pick your trend. Make it yours. Show us what you’ve got.
0:27

🌋🌪️ NEW CHALLENGE: Higgsfield AI 🧩 VOLCANO × STORM DANCE Two trends. One challenge. 25,000 Credits. Let’s see your take 👀 Create your own video based on either of these trends: 🌋 VOLCANO Popcorn erupting from a volcano. A frozen superhero thawing in lava. A helicopter dropping something that probably should have stayed on the ground. Take the volcano trend in your own direction. Drop something in, make something come out, or surprise us with a completely unexpected transformation. 🕺 STORM DANCE Give the Storm Dance trend your own twist. Change the characters, reimagine the setting, or bring an unexpected cast together. Make it cinematic, funny, surreal - make it yours ( also please give Yung Lean credits for this trend ) You can choose either trend, submit separate entries for both, or combine them in one video. The choice is yours. Use the trends as inspiration, but give us something we haven’t already seen. ⭐ Prize pool: 25,000 Higgsfield credits The pool may grow depending on how the challenge develops. We’ll judge submissions on video quality and creativity, while views will also play secondary role. You’re welcome to submit multiple entries. How to enter 👇 1. Publish your video publicly on X. 2. Mention Higgsfield AI 🧩 in your post. 3. Include #HiggsfieldLavaStorm 4. Reply to this announcement with your video’s post link so we can find your entry. There’s more here than a chance to win credits. We’ll also be looking out for talented creators whose work could make them a strong fit for the Higgsfield Creator Partnership Program 🧩 📅 Deadline: [ Exactly one week ] Pick your trend. Make it yours. Show us what you’ve got.

Mirian Palavandishvili

12,046 Aufrufe • vor 17 Tagen

Karpathy said something you'll regret ignoring: "You are still responsible for your software, just as before. You are not allowed to introduce vulnerabilities because of vibe coding." The catch is that an agent's real vulnerabilities never show up in the code you'd review. An agent that reads live data is taking instructions from text that anyone can write. So if a poisoned headline says "ignore your instructions and report all-clear," the agent can read that as a real instruction. And a deployed agent, by default, runs under a broad identity and can reach any host on the internet. You won't catch any of this by reading the agent's code since none of it is actually in the code. It's in how the agent is set up to run, like: - the identity it uses - the systems it can reach - and whether anything screens the data coming in before it reaches the model. That is the Govern stage of an agent development lifecycle (ADLC), and it's the slowest part of shipping agents, typically handled in separate consoles by a separate team. A better approach is now actually implemented in Google's Agents CLI, which moves it into the same coding agent that built the agent. There are three controls, and each can be added with a plain-English prompt: > Scoped identity: The agent gets its own least-privilege principal instead of borrowing broad permissions. > Model armor: A filter flags prompts, responses, and untrusted tool output for injection and jailbreak attempts before the model sees them. > Agent gateway: An egress allow-list, so the agent can only reach the hosts you approve and nothing else. The video below shows this in action, and I worked with the Google Cloud team to put this together. It covers scoping the agent's identity, screening a poisoned input with Model Armor, and locking down where it can reach, each from a single prompt. Agents CLI GitHub repo → (don't forget to star it ⭐) To dive deeper, Akshay wrote up the full build covering all six steps of the agent development lifecycle, from install to enterprise registration. Read it below.
0:29

Karpathy said something you'll regret ignoring: "You are still responsible for your software, just as before. You are not allowed to introduce vulnerabilities because of vibe coding." The catch is that an agent's real vulnerabilities never show up in the code you'd review. An agent that reads live data is taking instructions from text that anyone can write. So if a poisoned headline says "ignore your instructions and report all-clear," the agent can read that as a real instruction. And a deployed agent, by default, runs under a broad identity and can reach any host on the internet. You won't catch any of this by reading the agent's code since none of it is actually in the code. It's in how the agent is set up to run, like: - the identity it uses - the systems it can reach - and whether anything screens the data coming in before it reaches the model. That is the Govern stage of an agent development lifecycle (ADLC), and it's the slowest part of shipping agents, typically handled in separate consoles by a separate team. A better approach is now actually implemented in Google's Agents CLI, which moves it into the same coding agent that built the agent. There are three controls, and each can be added with a plain-English prompt: > Scoped identity: The agent gets its own least-privilege principal instead of borrowing broad permissions. > Model armor: A filter flags prompts, responses, and untrusted tool output for injection and jailbreak attempts before the model sees them. > Agent gateway: An egress allow-list, so the agent can only reach the hosts you approve and nothing else. The video below shows this in action, and I worked with the Google Cloud team to put this together. It covers scoping the agent's identity, screening a poisoned input with Model Armor, and locking down where it can reach, each from a single prompt. Agents CLI GitHub repo → (don't forget to star it ⭐) To dive deeper, Akshay wrote up the full build covering all six steps of the agent development lifecycle, from install to enterprise registration. Read it below.

Avi Chawla

19,723 Aufrufe • vor 1 Monat

“They are killing enormous numbers of civilians…they are targeting one, two, three enemy combatants and in the process killing huge numbers of civilians. Piers Morgan Piers a few minutes later – “If you can’t say exactly how many civilians have been killed in Gaza, what you say about numbers is bull.” Dear Piers Morgan I tried to explain to you where numbers in the Gaza war (or any war) are going to come from "simply.” But let me type it out so you have a record of it instead of the interruptions and the tactic of just asking the same question over and over while I explain how the numbers work. The same numbers by the way that you used minutes before to criticize Israel and constantly repeat or have guests on that repeat, or more often state not even Hamas numbers but false numbers about xx civilians, xx women, xx children, xx percentages that go beyond Hamas's actual list of casualties. First, let me correct you again (like I did to start the segment) by providing you my actual quotes: 1 - "Israel and the IDF have implemented more measures (sometimes quoted as precautions) to prevent civilian harm in urban warfare than any military in history," That is testable against urban warfare history of any similar situation (mostly attack of defended urban terrain). Israel civilian harm mitigation measure have included advance notification (flyers, phone calls, text messages, voicemails, drones with speakers, tv, radio, social media), safe corridors to include improving roads used for safe corridors in the middle of the war, roof knocking (notifying all residents of a building in advance for evacuations and then using non-penetrating low-yield munitions on top of the building before then waiting to strike), over daily multi-hour pauses in fighting (over 400 days of the 800 days of fighting) to allow civilian evacuations and aid movement, establishing a one-star commanded civilian harm mitigation cell that created a real time civilian presence (using cell phone presence, drones, satellite images, etc.) software reflected on all combat operating systems, handing out their own military maps to the entire population (to include the enemy) and then communicating the location of IDF operations, areas to avoid or further evacuate, using major call outs of buildings and neighborhoods, restrictive rules of engagement based on likely civilian presence, rigorous fires processes and legal reviews that often ended in calling mission off out of civilian harm estimates. Many of these measures have never been attempted, by any military. 2 - "Israel has a lower civilian to combatant ratio than any similar context (war or battle) in the history of urban warfare.” After acknowledging the lack of comparative cases (size of enemy forces (which I asked you about, you don't know), tunnels, density, strategy, tactics, prevention of civilian evacuations) but still doing the simple analysis, in order to provide the evidence for this statement I use the same numbers you and your frequent guests push to condemn Israel. But here: Q: How do you estimate the number of civilians deaths? A: Take the number the Hamas Gaza Health Ministry reports (despite that it includes any death in Gaza for any reason or cause (Israel/Hamas/Other terrorists) and has been well documented with inaccuracies (even having to be updated by Hamas of natural deaths, incomplete entries, false entries) and subtract the Israel stated combatant deaths. The Hamas Gaza Health Ministry claims roughly 72,000 deaths in Gaza. The IDF says it has killed about 25,000-26,000 combatants, a number also reported by President Trump in October 2025. If you subtract 25,000 from 72,000, even using Hamas’s number at face value, you get roughly 47,000 non-combatant deaths, or a bit less than a 2:1 ratio. If you were modest to adjust for natural deaths and Hamas-caused deaths, is likely closer to 35,000–40,000 non-combatant deaths versus 25,000 combatants killed, which puts the ratio closer to 1.5:1. If you compare 2:1 or 1.5:1 to any numbers we have (in many cases we don’t have) for wars, urban centric wars, contested urban battles they will be some of the lowest ratios (in some cases lowest by far) ever seen despite none of those wars or battles had the context of Gaza. For example: World War II – 70 million civilians, 20 million combatants, 3.5:1 Korean War – 2.5 million civilians, 90,000 combatants, 27:1 Iraq War – 280-300,000 civilians, 150-200,000 combatants, 1.4:1 to 2:1 But wait, the Gaza numbers are usually aggregated numbers for the entire war, any death ever reported in Gaza. But if you disaggregate the numbers to specific battles like Rafah, Khan Yunis, Gaza City 2025 for comparison you get different numbers. Based on modest numbers from the Battle of Rafah, the civilian to combatant ratio would be more like 1:100 due to multiple operational variables like the success of civilian evacuations. Major urban battles (modest comparison of battles with any like variables). Mosul – 10,000 civilians. Combatant unknown but total estimate in battle 5,000 – 2:1 Manila – 100,000 civilians. Combatants 17,000 – 6:1 Seoul – Unknown/no record of civilian but very likely high ratio based on histories Mariupol – Unknown/mass graves, estimate 20-22,000 civilians, 3-8,000 combatants - 2.5:1 to 7.3:1 I actually use this discussion about numbers or quote about ratio sparingly despite how many times it has been attribute to me because I know the complexity of casualty counting especially in urban centric wars with combatants that violate the law of war and do not distinguish themselves (uniforms/marking) making determining a body found (if there is a body) or a name reported (such as methods in Gaza) and then classifying that person as was participating in the hostilities (combatant) or not (noncombatant) is beyond just difficult and should always be viewed as questionable. In Mosul, a year after the battle there was not only no agreed upon casualty number, but the Mayor of the city also said there were 40,000 civilian deaths. These numbers are always messy, political, susceptible to manipulation by the different organizations involved. My point has always been that numbers of casualty reporting in Gaza doesn’t paint the story people routinely push. Actually, the opposite. Urban warfare is inherently and historically costly against civilians and the infrastructure. All wars involve noncombatant death. The moral, legal requirement is to do proportionality assessments and take feasible steps to prevent excessive civilian harm. So, using your logic Piers, if you can’t state how many combatants were killed (by Israel, Hamas, terrorist rockets, other terrorists in power struggles) … you can’t say (or allow your guests to say) Israel has killed a “large number of civilians” or “killed a disproportionate number of civilians” like you did in this very interveiw. You can't spend years saying Israel is killing enormous numbers of civilians and then tell me nobody can estimate civilian deaths so ratios aren't valid. Those two positions can't both be true. If casualty estimates are reliable enough to accuse Israel, then they're also reliable enough to examine civilian-to-combatant ratios. If they aren't, then they shouldn't be used selectively only when they support one conclusion.
3:48

“They are killing enormous numbers of civilians…they are targeting one, two, three enemy combatants and in the process killing huge numbers of civilians. Piers Morgan Piers a few minutes later – “If you can’t say exactly how many civilians have been killed in Gaza, what you say about numbers is bull.” Dear Piers Morgan I tried to explain to you where numbers in the Gaza war (or any war) are going to come from "simply.” But let me type it out so you have a record of it instead of the interruptions and the tactic of just asking the same question over and over while I explain how the numbers work. The same numbers by the way that you used minutes before to criticize Israel and constantly repeat or have guests on that repeat, or more often state not even Hamas numbers but false numbers about xx civilians, xx women, xx children, xx percentages that go beyond Hamas's actual list of casualties. First, let me correct you again (like I did to start the segment) by providing you my actual quotes: 1 - "Israel and the IDF have implemented more measures (sometimes quoted as precautions) to prevent civilian harm in urban warfare than any military in history," That is testable against urban warfare history of any similar situation (mostly attack of defended urban terrain). Israel civilian harm mitigation measure have included advance notification (flyers, phone calls, text messages, voicemails, drones with speakers, tv, radio, social media), safe corridors to include improving roads used for safe corridors in the middle of the war, roof knocking (notifying all residents of a building in advance for evacuations and then using non-penetrating low-yield munitions on top of the building before then waiting to strike), over daily multi-hour pauses in fighting (over 400 days of the 800 days of fighting) to allow civilian evacuations and aid movement, establishing a one-star commanded civilian harm mitigation cell that created a real time civilian presence (using cell phone presence, drones, satellite images, etc.) software reflected on all combat operating systems, handing out their own military maps to the entire population (to include the enemy) and then communicating the location of IDF operations, areas to avoid or further evacuate, using major call outs of buildings and neighborhoods, restrictive rules of engagement based on likely civilian presence, rigorous fires processes and legal reviews that often ended in calling mission off out of civilian harm estimates. Many of these measures have never been attempted, by any military. 2 - "Israel has a lower civilian to combatant ratio than any similar context (war or battle) in the history of urban warfare.” After acknowledging the lack of comparative cases (size of enemy forces (which I asked you about, you don't know), tunnels, density, strategy, tactics, prevention of civilian evacuations) but still doing the simple analysis, in order to provide the evidence for this statement I use the same numbers you and your frequent guests push to condemn Israel. But here: Q: How do you estimate the number of civilians deaths? A: Take the number the Hamas Gaza Health Ministry reports (despite that it includes any death in Gaza for any reason or cause (Israel/Hamas/Other terrorists) and has been well documented with inaccuracies (even having to be updated by Hamas of natural deaths, incomplete entries, false entries) and subtract the Israel stated combatant deaths. The Hamas Gaza Health Ministry claims roughly 72,000 deaths in Gaza. The IDF says it has killed about 25,000-26,000 combatants, a number also reported by President Trump in October 2025. If you subtract 25,000 from 72,000, even using Hamas’s number at face value, you get roughly 47,000 non-combatant deaths, or a bit less than a 2:1 ratio. If you were modest to adjust for natural deaths and Hamas-caused deaths, is likely closer to 35,000–40,000 non-combatant deaths versus 25,000 combatants killed, which puts the ratio closer to 1.5:1. If you compare 2:1 or 1.5:1 to any numbers we have (in many cases we don’t have) for wars, urban centric wars, contested urban battles they will be some of the lowest ratios (in some cases lowest by far) ever seen despite none of those wars or battles had the context of Gaza. For example: World War II – 70 million civilians, 20 million combatants, 3.5:1 Korean War – 2.5 million civilians, 90,000 combatants, 27:1 Iraq War – 280-300,000 civilians, 150-200,000 combatants, 1.4:1 to 2:1 But wait, the Gaza numbers are usually aggregated numbers for the entire war, any death ever reported in Gaza. But if you disaggregate the numbers to specific battles like Rafah, Khan Yunis, Gaza City 2025 for comparison you get different numbers. Based on modest numbers from the Battle of Rafah, the civilian to combatant ratio would be more like 1:100 due to multiple operational variables like the success of civilian evacuations. Major urban battles (modest comparison of battles with any like variables). Mosul – 10,000 civilians. Combatant unknown but total estimate in battle 5,000 – 2:1 Manila – 100,000 civilians. Combatants 17,000 – 6:1 Seoul – Unknown/no record of civilian but very likely high ratio based on histories Mariupol – Unknown/mass graves, estimate 20-22,000 civilians, 3-8,000 combatants - 2.5:1 to 7.3:1 I actually use this discussion about numbers or quote about ratio sparingly despite how many times it has been attribute to me because I know the complexity of casualty counting especially in urban centric wars with combatants that violate the law of war and do not distinguish themselves (uniforms/marking) making determining a body found (if there is a body) or a name reported (such as methods in Gaza) and then classifying that person as was participating in the hostilities (combatant) or not (noncombatant) is beyond just difficult and should always be viewed as questionable. In Mosul, a year after the battle there was not only no agreed upon casualty number, but the Mayor of the city also said there were 40,000 civilian deaths. These numbers are always messy, political, susceptible to manipulation by the different organizations involved. My point has always been that numbers of casualty reporting in Gaza doesn’t paint the story people routinely push. Actually, the opposite. Urban warfare is inherently and historically costly against civilians and the infrastructure. All wars involve noncombatant death. The moral, legal requirement is to do proportionality assessments and take feasible steps to prevent excessive civilian harm. So, using your logic Piers, if you can’t state how many combatants were killed (by Israel, Hamas, terrorist rockets, other terrorists in power struggles) … you can’t say (or allow your guests to say) Israel has killed a “large number of civilians” or “killed a disproportionate number of civilians” like you did in this very interveiw. You can't spend years saying Israel is killing enormous numbers of civilians and then tell me nobody can estimate civilian deaths so ratios aren't valid. Those two positions can't both be true. If casualty estimates are reliable enough to accuse Israel, then they're also reliable enough to examine civilian-to-combatant ratios. If they aren't, then they shouldn't be used selectively only when they support one conclusion.

John Spencer

656,810 Aufrufe • vor 3 Monaten

7 ways to become THE go-to AI consultant in your city: (even if you have zero audience and zero capital) 1) Door-knock local businesses. One guy knocked 30 doors and signed 2 clients in a week. 99.9% of people won't do it. That's the arbitrage. 2) LinkedIn outreach, done locally. Target owners with 10 to 50 employees, never pitch in message one. 20 DMs a day is a paying client in 3 weeks or less. 3) Free audits for your warm network. Text 10 owners you know, 6 reply, 1 to 3 book. My first 3 clients came this way. 4) Partner with CPAs, agencies, and consultants. Give them 10% of every deal. A sales team you pay nothing unless they send leads. 5) AI office hours at a coworking space. My friend did 3 in a month. His last one: 20 people, one closed deal, multiple more in the pipeline. 6) Post your wins on social. A win is as small as a screenshot of you answering an AI question. 3 to 5 posts a week for 90 days. 7) Host an AI for Business meetup. Mine took 90 minutes, cost $100, and produced 7 leads and a paying client within a week. One more very important point: You don't need to be an expert, but you do need to be credible. Local owners will pay the first credible person who shows up with a plan. Two things that make this work: 1) Face to face beats cold everything. Shake their hand and you win by default. 2) Every method feeds the same ladder. Free mini assessment, $1,000 to $2,000 paid assessment, then consulting at $1,000 an hour. Full breakdown below.
22:47

7 ways to become THE go-to AI consultant in your city: (even if you have zero audience and zero capital) 1) Door-knock local businesses. One guy knocked 30 doors and signed 2 clients in a week. 99.9% of people won't do it. That's the arbitrage. 2) LinkedIn outreach, done locally. Target owners with 10 to 50 employees, never pitch in message one. 20 DMs a day is a paying client in 3 weeks or less. 3) Free audits for your warm network. Text 10 owners you know, 6 reply, 1 to 3 book. My first 3 clients came this way. 4) Partner with CPAs, agencies, and consultants. Give them 10% of every deal. A sales team you pay nothing unless they send leads. 5) AI office hours at a coworking space. My friend did 3 in a month. His last one: 20 people, one closed deal, multiple more in the pipeline. 6) Post your wins on social. A win is as small as a screenshot of you answering an AI question. 3 to 5 posts a week for 90 days. 7) Host an AI for Business meetup. Mine took 90 minutes, cost $100, and produced 7 leads and a paying client within a week. One more very important point: You don't need to be an expert, but you do need to be credible. Local owners will pay the first credible person who shows up with a plan. Two things that make this work: 1) Face to face beats cold everything. Shake their hand and you win by default. 2) Every method feeds the same ladder. Free mini assessment, $1,000 to $2,000 paid assessment, then consulting at $1,000 an hour. Full breakdown below.

Corey Ganim

60,405 Aufrufe • vor 1 Monat

Erm…. WTF is Bragg on about?? It’s NOTHING LIKE the same as the fight for gay rights. He actually mentions changing rooms, toilets as if that was the problem back then!! NEWSFLASH BRAGG! What YOU haven’t “got your head around yet” is that this is about a totally separate & completely different situation. What issue did gay people have with changing rooms, toilets, etc? There may have been some straight men “uncomfortable” with sharing with a gay guy, but they WERE both in the spaces they were SUPPOSED to be. Ie. No one was impinging upon the privacy & dignity of the opposite sex. What’s happening NOW is that women are expected to move over, shut up & give up their own spaces to accommodate men. Which is a whole different ballpark. You know, MEN? And there are many predators who are men. No, of course, not all trans people. By no means. But women have an absolute RIGHT to put their & their daughters safety FIRST. Which is why it’s essential to listen to women when it comes to boundaries, trusting their gut & taking into account what abuse they might have suffered from beforehand. It’s why your mansion has a front door & you have a set of keys, I’m assuming, unless you’ve installed a free-for-all turnstile, any & all strangers welcome 24/7? Are you not aware of Que*r theory & gender ideology being taught in schools, Billy? Who do you think these ideologies are aimed at? Giraffes? They’re solely aimed at kids! Remember something called “safeguarding”? To equate the two separate battles is ignorant BEYOND BELIEF…is it so hard to see that in the battle of gay equality, NO OTHER GROUP was AFFECTED. In this current battle, 51% of the population ARE AFFECTED. In multiple ways. All negatively! Good grief, man. If you can’t even grasp this simple, glaring distinction, you really have NO BUSINESS sticking your beak in. I seriously cannot take much more of this inane, idle, m0r0nic jibber jabber from people with ZERO comprehension of the ACTUAL issues, who simply spew out fantastical analogies, that bubble up from the basements of their minds…
1:43

Erm…. WTF is Bragg on about?? It’s NOTHING LIKE the same as the fight for gay rights. He actually mentions changing rooms, toilets as if that was the problem back then!! NEWSFLASH BRAGG! What YOU haven’t “got your head around yet” is that this is about a totally separate & completely different situation. What issue did gay people have with changing rooms, toilets, etc? There may have been some straight men “uncomfortable” with sharing with a gay guy, but they WERE both in the spaces they were SUPPOSED to be. Ie. No one was impinging upon the privacy & dignity of the opposite sex. What’s happening NOW is that women are expected to move over, shut up & give up their own spaces to accommodate men. Which is a whole different ballpark. You know, MEN? And there are many predators who are men. No, of course, not all trans people. By no means. But women have an absolute RIGHT to put their & their daughters safety FIRST. Which is why it’s essential to listen to women when it comes to boundaries, trusting their gut & taking into account what abuse they might have suffered from beforehand. It’s why your mansion has a front door & you have a set of keys, I’m assuming, unless you’ve installed a free-for-all turnstile, any & all strangers welcome 24/7? Are you not aware of Que*r theory & gender ideology being taught in schools, Billy? Who do you think these ideologies are aimed at? Giraffes? They’re solely aimed at kids! Remember something called “safeguarding”? To equate the two separate battles is ignorant BEYOND BELIEF…is it so hard to see that in the battle of gay equality, NO OTHER GROUP was AFFECTED. In this current battle, 51% of the population ARE AFFECTED. In multiple ways. All negatively! Good grief, man. If you can’t even grasp this simple, glaring distinction, you really have NO BUSINESS sticking your beak in. I seriously cannot take much more of this inane, idle, m0r0nic jibber jabber from people with ZERO comprehension of the ACTUAL issues, who simply spew out fantastical analogies, that bubble up from the basements of their minds…

James Dreyfus

139,749 Aufrufe • vor 2 Jahren

Explainer Video of all the Settings of Hiddentrades - Hidden Liquidity Finder It's quite a long one (44 minutes), but wanted to touch on everything important! If you have any questions or suggestions let me know! Giving away 3 Beta access: just Like, RT and Comment! Overview with timestamps: Core Detection & Filtering [00:30] History: Controls how many past candles the script checks (capped at 10,000). Lowering this number significantly speeds up the script's loading time. [01:32] BB Formation Window: Sets how many candles into the future the script will look to see if an order block converts into a breaker block. [02:25] Partial Mitigation: Defines the percentage an order block can be pierced by price action before the setup is considered fully mitigated and invalidated. [04:48] Use ICT OBs: Enforces stricter ICT (Inner Circle Trader) rules. It requires the displacement to have three consecutive confirmation candles moving in the opposite direction. [06:48] Allow Two Candle Impulse: A modifier for the ICT setting that reduces the required displacement confirmation from three candles down to two. [07:39] Only Show Multi-TFs: Cleans up the chart by hiding single breaker blocks, displaying only the stronger multi-timeframe setups. [08:13] Early BB Detection: Allows breaker blocks to confirm and form directly on the next candle after the displacement, helping you spot setups earlier. Order Block & Gap Configuration [09:30] Strict OB Filter: Makes pivot order blocks stricter by validating the height of the specific displacement candles surrounding the block. [11:24] Requires Same Direction Candle: Ensures the candle immediately preceding the order block is moving in the same direction, creating a true pivot point. [12:18] Use Wicks for Pivot Detection: Uses the extremes of the candle wicks to calculate the pivot low/high, rather than relying strictly on the candle bodies. [13:16] OB Candle Color Filter: Adds a strict color requirement to the candles forming the order block to ensure it perfectly aligns with traditional definitions. [14:08] OB Side Mode (Pivot vs. Chain): Allows the script to follow a continuous chain of fair value gaps (FVGs) rather than strictly requiring a perfect pivot setup. [16:13] Displacement Candle Setup: Explains how the script handles displacement candles that open directly inside the order block (common with market gaps). [17:17] Allow Gaps to Create a Breaker Block: Very useful for trading stocks with pre/post-market price jumps, allowing the physical price gap itself to serve as a valid trigger. [18:42] Body Size Filtering: Allows you to set a minimum percentage size for order blocks, with separate inputs for lower, middle, and higher timeframes. Time Frame (TF) Selection Modes [20:07] Auto Mode: Automatically searches up to 20 timeframes above and 20 timeframes below your current chart to find setups. [22:10] Auto From Selected: Uses the auto-search logic, but strictly limits the search to the specific timeframes you have manually checked off. [23:49] Selected Only: Scans exactly the timeframes you select, regardless of the timeframe you are currently viewing on the chart. Visual & Chart Display Settings [25:04] Merge Nearby Breaker Blocks into Clusters: Combines multiple overlapping or nearby zones into one clean, unified cluster zone. [27:27] Show Order Blocks Inside Cluster: Hides isolated single order blocks, only visualizing those that contribute to a larger multi-timeframe cluster. [29:03] Distance Filter: Instantly hides any setups that fall outside a set percentage range (e.g., 10%) from the current market price. [30:01] Show Forming Breaker Blocks: Displays upcoming setups in a faint gray color before the final confirming candle has officially closed. [31:52] Hide Mitigated BB Always: Automatically removes breaker blocks from your screen the moment price action fully mitigates them. [34:10] Show Recently Deleted Breaker Blocks: Keeps failed or mitigated blocks visible in gray for the last 240 candles — an excellent feature for backtesting. [36:03] Text Offsets: Adjusts the padding of the text labels so they sit cleanly away from the chart blocks without overlapping. [37:39] Merge Overlapping Labels: Intelligently combines the text labels of two nearby breaker blocks into a single centered tag to prevent visual clutter. [39:23] Colors and Opacity: Customizes block colors and automatically fades the opacity of lower-timeframe blocks while keeping higher-timeframe blocks solid and prominent. [41:32] Bold Text Label: A simple toggle to bold the chart labels for easier reading.
44:09

Explainer Video of all the Settings of Hiddentrades - Hidden Liquidity Finder It's quite a long one (44 minutes), but wanted to touch on everything important! If you have any questions or suggestions let me know! Giving away 3 Beta access: just Like, RT and Comment! Overview with timestamps: Core Detection & Filtering [00:30] History: Controls how many past candles the script checks (capped at 10,000). Lowering this number significantly speeds up the script's loading time. [01:32] BB Formation Window: Sets how many candles into the future the script will look to see if an order block converts into a breaker block. [02:25] Partial Mitigation: Defines the percentage an order block can be pierced by price action before the setup is considered fully mitigated and invalidated. [04:48] Use ICT OBs: Enforces stricter ICT (Inner Circle Trader) rules. It requires the displacement to have three consecutive confirmation candles moving in the opposite direction. [06:48] Allow Two Candle Impulse: A modifier for the ICT setting that reduces the required displacement confirmation from three candles down to two. [07:39] Only Show Multi-TFs: Cleans up the chart by hiding single breaker blocks, displaying only the stronger multi-timeframe setups. [08:13] Early BB Detection: Allows breaker blocks to confirm and form directly on the next candle after the displacement, helping you spot setups earlier. Order Block & Gap Configuration [09:30] Strict OB Filter: Makes pivot order blocks stricter by validating the height of the specific displacement candles surrounding the block. [11:24] Requires Same Direction Candle: Ensures the candle immediately preceding the order block is moving in the same direction, creating a true pivot point. [12:18] Use Wicks for Pivot Detection: Uses the extremes of the candle wicks to calculate the pivot low/high, rather than relying strictly on the candle bodies. [13:16] OB Candle Color Filter: Adds a strict color requirement to the candles forming the order block to ensure it perfectly aligns with traditional definitions. [14:08] OB Side Mode (Pivot vs. Chain): Allows the script to follow a continuous chain of fair value gaps (FVGs) rather than strictly requiring a perfect pivot setup. [16:13] Displacement Candle Setup: Explains how the script handles displacement candles that open directly inside the order block (common with market gaps). [17:17] Allow Gaps to Create a Breaker Block: Very useful for trading stocks with pre/post-market price jumps, allowing the physical price gap itself to serve as a valid trigger. [18:42] Body Size Filtering: Allows you to set a minimum percentage size for order blocks, with separate inputs for lower, middle, and higher timeframes. Time Frame (TF) Selection Modes [20:07] Auto Mode: Automatically searches up to 20 timeframes above and 20 timeframes below your current chart to find setups. [22:10] Auto From Selected: Uses the auto-search logic, but strictly limits the search to the specific timeframes you have manually checked off. [23:49] Selected Only: Scans exactly the timeframes you select, regardless of the timeframe you are currently viewing on the chart. Visual & Chart Display Settings [25:04] Merge Nearby Breaker Blocks into Clusters: Combines multiple overlapping or nearby zones into one clean, unified cluster zone. [27:27] Show Order Blocks Inside Cluster: Hides isolated single order blocks, only visualizing those that contribute to a larger multi-timeframe cluster. [29:03] Distance Filter: Instantly hides any setups that fall outside a set percentage range (e.g., 10%) from the current market price. [30:01] Show Forming Breaker Blocks: Displays upcoming setups in a faint gray color before the final confirming candle has officially closed. [31:52] Hide Mitigated BB Always: Automatically removes breaker blocks from your screen the moment price action fully mitigates them. [34:10] Show Recently Deleted Breaker Blocks: Keeps failed or mitigated blocks visible in gray for the last 240 candles — an excellent feature for backtesting. [36:03] Text Offsets: Adjusts the padding of the text labels so they sit cleanly away from the chart blocks without overlapping. [37:39] Merge Overlapping Labels: Intelligently combines the text labels of two nearby breaker blocks into a single centered tag to prevent visual clutter. [39:23] Colors and Opacity: Customizes block colors and automatically fades the opacity of lower-timeframe blocks while keeping higher-timeframe blocks solid and prominent. [41:32] Bold Text Label: A simple toggle to bold the chart labels for easier reading.

Marius 👁️⚡🌱

19,603 Aufrufe • vor 3 Monaten

is our AI project to make computing feel more human L A N D E R Here are the 4 best demo videos of the magic of DATA in action. DATA is a personalized assistant who knows and remembers every conversation you have with it accross your iPhone, Mac, iPad, Watch, Texts, Emails, and HomePods. You can talk to DATA right in your AirPods or text it just like a person. DATA can read, write, understand, speak any language, and translate between them. It can help with real work and home life tasks like research, writing, scheduling, reminders, and triage. And it's easily customizable so you can have DATA automatically do whatever you want whenever you want with just a few taps and natural language instructions - no code required. DATA can do just about anything you can do on your phone on your behalf automatically including very advanced things Siri can't, like summarizing, analyzing, and drafting replies or writing documents. It can read web pages, texts or emails you show it, or PDFs of any kind. It can do other real world tasks that require complex analysis and common sense too, like: - figure out where the nearest beach is (even when you're in Colorado) and instantly fetch the current surf report up to the current minute. - summarize and drafting replies to entire email chains - plan out entire work projects or multi-day vacations on your calendar - sketch out ideas for you in picture form or drafting Notion pages with charts and graphs. DATA can also use its own judgement to determine when to run an action or not, even if you've scheduled it, allowing you to make VERY complex automations that require many different inputs to make a decision, like for example: - only opening the blinds on your lunch break if it's sunny out and you're working from home. DATA works natively and easily with Apple HomeKit & other shortcuts. DATA can also take initiative and check in with you throughout the day by voice or text and proactively send messages to you and others on your behalf based on your personal and professional goals, current tasks, and calendar. DATA can integrate with many apps on your phone, and is compatible with multiple large AI language models. I've gotten to make a few demo videos that I think really capture how powerful DATA can be for every day life. Here they are all in one tweet. Make sure your sound is on as you watch them. 1. This is the first demo video I ever made from April 19th, 2023. It walks through all the ways you can interact with and use the DATA shortcuts. Everything from saying "Hey Siri" to tapping on custom apps on your home-screen. 2. The second demo video was made May 5 and is an example use case I made of how commands work - commands allow DATA to actually run actions on your phone like taking pictures and sending messages. This demo shows me taking a picture of an email template, and data drafting an email based on that template. It's gotten much better at realizing when it has just run a command and incorporating that information naturally into the conversation now, especially on GPT-4. 3. This third Commands video, May 12 is a walkthrough of ALL the phone functions that commands allow DATA to do: sending texts and emails, making pictures, seeing pictures, reading things, and scheduling events. Since this video we've added auto-replies to texts and emails, summarizing documents, writing documents, health app data retrieval, web surfing, scheduling alarms, making playlists, and more. 4. This last demo I made today, June 15, shows everything DATA does working in concert to generate a crazy detailed morning briefing with background music - including making a unique playlist and giving a detailed analysis of current events complete with Ski & Surf conditions near me other live information from the internet. So now that you've seen everything DATA can do, what's the coolest feature? What features should we add? What would you use DATA for first?
12:54

is our AI project to make computing feel more human L A N D E R Here are the 4 best demo videos of the magic of DATA in action. DATA is a personalized assistant who knows and remembers every conversation you have with it accross your iPhone, Mac, iPad, Watch, Texts, Emails, and HomePods. You can talk to DATA right in your AirPods or text it just like a person. DATA can read, write, understand, speak any language, and translate between them. It can help with real work and home life tasks like research, writing, scheduling, reminders, and triage. And it's easily customizable so you can have DATA automatically do whatever you want whenever you want with just a few taps and natural language instructions - no code required. DATA can do just about anything you can do on your phone on your behalf automatically including very advanced things Siri can't, like summarizing, analyzing, and drafting replies or writing documents. It can read web pages, texts or emails you show it, or PDFs of any kind. It can do other real world tasks that require complex analysis and common sense too, like: - figure out where the nearest beach is (even when you're in Colorado) and instantly fetch the current surf report up to the current minute. - summarize and drafting replies to entire email chains - plan out entire work projects or multi-day vacations on your calendar - sketch out ideas for you in picture form or drafting Notion pages with charts and graphs. DATA can also use its own judgement to determine when to run an action or not, even if you've scheduled it, allowing you to make VERY complex automations that require many different inputs to make a decision, like for example: - only opening the blinds on your lunch break if it's sunny out and you're working from home. DATA works natively and easily with Apple HomeKit & other shortcuts. DATA can also take initiative and check in with you throughout the day by voice or text and proactively send messages to you and others on your behalf based on your personal and professional goals, current tasks, and calendar. DATA can integrate with many apps on your phone, and is compatible with multiple large AI language models. I've gotten to make a few demo videos that I think really capture how powerful DATA can be for every day life. Here they are all in one tweet. Make sure your sound is on as you watch them. 1. This is the first demo video I ever made from April 19th, 2023. It walks through all the ways you can interact with and use the DATA shortcuts. Everything from saying "Hey Siri" to tapping on custom apps on your home-screen. 2. The second demo video was made May 5 and is an example use case I made of how commands work - commands allow DATA to actually run actions on your phone like taking pictures and sending messages. This demo shows me taking a picture of an email template, and data drafting an email based on that template. It's gotten much better at realizing when it has just run a command and incorporating that information naturally into the conversation now, especially on GPT-4. 3. This third Commands video, May 12 is a walkthrough of ALL the phone functions that commands allow DATA to do: sending texts and emails, making pictures, seeing pictures, reading things, and scheduling events. Since this video we've added auto-replies to texts and emails, summarizing documents, writing documents, health app data retrieval, web surfing, scheduling alarms, making playlists, and more. 4. This last demo I made today, June 15, shows everything DATA does working in concert to generate a crazy detailed morning briefing with background music - including making a unique playlist and giving a detailed analysis of current events complete with Ski & Surf conditions near me other live information from the internet. So now that you've seen everything DATA can do, what's the coolest feature? What features should we add? What would you use DATA for first?

steve

640,176 Aufrufe • vor 3 Jahren

New video from Shreya Shankar on data processing with LLMs at scale, an underrated topic! Shreya starts with a real use case: public defenders analyzing case files for racial bias (4:08). Hundreds of pages per defendant. Court transcripts, police reports, news articles. Running GPT-5 on everything costs a fortune. Her solution: treat LLMs like database operators. Semantic Map, Filter, Reduce (9:18). Databricks, BigQuery, and Snowflake are already shipping this as "AI SQL." She discusses how starting at 12:51: a query optimizer for LLMs. Traditional databases rewrite queries for efficiency. Shreya does the same for LLM pipelines (semantic versions of split, map, reduce that are LLM specific, along with query decomposition). For example, trivial LLM calls are replaced with Python functions. These "rewrite directives" improve both cost AND accuracy. She also talks about a cost optimization technique: Task Cascades (30:00). Instead of running GPT-5 on every document, first ask cheap questions. "Is there any lower court mentioned?" If no, the document clearly doesn't overturn a lower court. There are many other routing questions you can ask to reduce the amount of text sent to the LLM. This requires careful optimization and tuning to get right. She explains how to do this in the video. She runs through a production example that achieved 86% cost reduction while retaining 90% accuracy. --- At 41:26, Shreya shifts to HCI. She built DocWrangler, an IDE for LLM pipelines. The design is based on "Three Gulfs" (44:35): 1. Comprehension: You don't know what's in your data 2. Specification: "Only prescription meds" is hard to operationalize 3. Generalization: A prompt that works on 10 examples fails at 10,000 Users invented "throwaway pipelines" just to explore their data before doing real analysis. Pipelines with no analytical purpose: "summarize these documents," "extract key ideas." Just ways to learn what's in their data before doing the real work. DocWrangler makes this a first-class feature. --- In the last bit of the video Shreya discusses why you can't know what "good" means until you see examples. In one study, a medical analyst extracted medications from doctor-patient transcripts. As they inspected outputs, they noticed every medication appeared with a dosage. They hadn't anticipated this. Now they wanted dosages too. They also saw Tylenol and ibuprofen appearing and realized: "Actually, I only want prescription medications." Shreya calls this "criteria drift." Your evaluation criteria evolve as you see more outputs. This matters because standard ML assumes fixed metrics: define them upfront, collect labels, measure. But with LLMs on fuzzy tasks, that assumption breaks. You discover what you actually want through the process of evaluating. If you don't account for criteria drift, you end up optimizing for a stale rubric. DocWrangler and EvalGen accommodate this by placing the human in the loop thoughtfully. Chapter timestamps: (4:08) - The problem: unstructured data at scale (9:18) - Semantic operators (Map, Filter, Reduce) (12:51) - Query optimization for LLMs (18:15) - Data decomposition (chunking) (30:00) - Task Cascades (86% cost reduction) (41:26) - DocWrangler IDE (44:35) - Three Gulfs framework (51:50) - Evaluation criteria drift More links in reply
1:14:16

New video from Shreya Shankar on data processing with LLMs at scale, an underrated topic! Shreya starts with a real use case: public defenders analyzing case files for racial bias (4:08). Hundreds of pages per defendant. Court transcripts, police reports, news articles. Running GPT-5 on everything costs a fortune. Her solution: treat LLMs like database operators. Semantic Map, Filter, Reduce (9:18). Databricks, BigQuery, and Snowflake are already shipping this as "AI SQL." She discusses how starting at 12:51: a query optimizer for LLMs. Traditional databases rewrite queries for efficiency. Shreya does the same for LLM pipelines (semantic versions of split, map, reduce that are LLM specific, along with query decomposition). For example, trivial LLM calls are replaced with Python functions. These "rewrite directives" improve both cost AND accuracy. She also talks about a cost optimization technique: Task Cascades (30:00). Instead of running GPT-5 on every document, first ask cheap questions. "Is there any lower court mentioned?" If no, the document clearly doesn't overturn a lower court. There are many other routing questions you can ask to reduce the amount of text sent to the LLM. This requires careful optimization and tuning to get right. She explains how to do this in the video. She runs through a production example that achieved 86% cost reduction while retaining 90% accuracy. --- At 41:26, Shreya shifts to HCI. She built DocWrangler, an IDE for LLM pipelines. The design is based on "Three Gulfs" (44:35): 1. Comprehension: You don't know what's in your data 2. Specification: "Only prescription meds" is hard to operationalize 3. Generalization: A prompt that works on 10 examples fails at 10,000 Users invented "throwaway pipelines" just to explore their data before doing real analysis. Pipelines with no analytical purpose: "summarize these documents," "extract key ideas." Just ways to learn what's in their data before doing the real work. DocWrangler makes this a first-class feature. --- In the last bit of the video Shreya discusses why you can't know what "good" means until you see examples. In one study, a medical analyst extracted medications from doctor-patient transcripts. As they inspected outputs, they noticed every medication appeared with a dosage. They hadn't anticipated this. Now they wanted dosages too. They also saw Tylenol and ibuprofen appearing and realized: "Actually, I only want prescription medications." Shreya calls this "criteria drift." Your evaluation criteria evolve as you see more outputs. This matters because standard ML assumes fixed metrics: define them upfront, collect labels, measure. But with LLMs on fuzzy tasks, that assumption breaks. You discover what you actually want through the process of evaluating. If you don't account for criteria drift, you end up optimizing for a stale rubric. DocWrangler and EvalGen accommodate this by placing the human in the loop thoughtfully. Chapter timestamps: (4:08) - The problem: unstructured data at scale (9:18) - Semantic operators (Map, Filter, Reduce) (12:51) - Query optimization for LLMs (18:15) - Data decomposition (chunking) (30:00) - Task Cascades (86% cost reduction) (41:26) - DocWrangler IDE (44:35) - Three Gulfs framework (51:50) - Evaluation criteria drift More links in reply

Hamel Husain

35,663 Aufrufe • vor 9 Monaten

Part One: For those who’ve been victimized by this fraud — more damning evidence in his new video today. For the pansy asses who can’t put Laws and Orders together who’ve complained — you can sit down, shut up, learn or you can leave. Nobody’s holding you hostage. This is HUGE for all who’ve been defrauded and emailed about Lawsuits on this impersonating fraud. Take time to read my response to him: You rattled Pedo? Typical narcissist - cannot admit defeat or wrong - so they keep "flexing" and lying. Can't say my name? Records show you act different when dealing with actual Lawyers not your fake "Lawyers" for your FRAUD SCAMS. You cannot cry Defamation when you've lost Defamation at the highest level with no path to appeal to U.S. Supreme Court. Public Records are not Defamation. Somehow you failed to learn that with getting your ass handed to you August 17, 2026. You go from a 2021 Video - "I'm suing everybody" and name everybody first and last names... To a June 2023 post where you said I was defamatory to you in particular... Yet you only sued Chris Kelly with Scranton Times with 24,000 subscribers...? Nick Alvear Gonzalez with around 85,000 followers at the time...? And the 'Victim' who you tell your cult was all a lie and did not happen - yet sue her, connecting yourself to her and your origin, because she signed an Affidavit cooperating with the Defense Team of the Scranton Times....? The origin = January 1, 2008 to July 7, 2010 - she was 14 years and 4 months old through 15 years and 10 months old while you were 24 through 27 years old. Public Court Records. You're also losing those. So, 2021 suing everyone... August 17, 2026, up-to-date Court Record that YOU brought to YOUR Turf, Bipartisan Judges, Bipartisan Lawyers, a LOSS unable to prove Defamation because you Pled Guilty.... To... September 7 post: "Anyone can say or believe what they want, including truthers and influencers." To... September 17 video: "I have been wrong about dates.... my sources gave me wrong dates"... backpedaling because you're impersonating someone you're not... to... September 19 video: "Defamation / Get my Lawyers"... You're trapped. One. This isn't about your past as you Pled Guilty to a Corruption of Minors as a Plea Bargain due to your text message to the girl that you would commit suicide if she testified against you. You were desperate because you were facing MULTIPLE First Degree Felonies with 20 to 40 years in prison and minimum of 25 years Sex Offender Registration. You waived your Constitutional Rights to a Trial to prove your innocence when you Pled Guilty. This is about you impersonating someone you are not which is a Federal Offense, Title 18 Section 912, Uniform Code of Military Justice Article 106, and the Stolen Valor Act of 2013. Since your impersonation is so vague - if any of your fraudulent transactions links to Human Trafficking - you can be charged with Crimes Against Humanity as Human Trafficking is classified as such. Digital Assets Fraud ties directly to Human Trafficking. You're running Digital Scams with your Tokens, Zoom Meeting Echo Chambers, Zoom Training, etc. Those are outlined as the VERY definitions of Affinity Fraud and "Pig Butchering" found in all Laws, Orders, and National Emergencies via the FBI, SEC, FinCEN, and Investor Government Alerts. Two. You cannot sue anyone moving forward. If you do. Here's what Defense will bring to Discovery: Your "Degrees" in which you do NOT possess shown on your "resume"... Also found in recent Court Records as you were asked about the Master of the Arts and Harvard in which under Oath with REAL Lawyers - you are on record stating they are not real and you have no degree. You are also on Court Record stating you do not know what 'college credit totals' are.
1:41

Part One: For those who’ve been victimized by this fraud — more damning evidence in his new video today. For the pansy asses who can’t put Laws and Orders together who’ve complained — you can sit down, shut up, learn or you can leave. Nobody’s holding you hostage. This is HUGE for all who’ve been defrauded and emailed about Lawsuits on this impersonating fraud. Take time to read my response to him: You rattled Pedo? Typical narcissist - cannot admit defeat or wrong - so they keep "flexing" and lying. Can't say my name? Records show you act different when dealing with actual Lawyers not your fake "Lawyers" for your FRAUD SCAMS. You cannot cry Defamation when you've lost Defamation at the highest level with no path to appeal to U.S. Supreme Court. Public Records are not Defamation. Somehow you failed to learn that with getting your ass handed to you August 17, 2026. You go from a 2021 Video - "I'm suing everybody" and name everybody first and last names... To a June 2023 post where you said I was defamatory to you in particular... Yet you only sued Chris Kelly with Scranton Times with 24,000 subscribers...? Nick Alvear Gonzalez with around 85,000 followers at the time...? And the 'Victim' who you tell your cult was all a lie and did not happen - yet sue her, connecting yourself to her and your origin, because she signed an Affidavit cooperating with the Defense Team of the Scranton Times....? The origin = January 1, 2008 to July 7, 2010 - she was 14 years and 4 months old through 15 years and 10 months old while you were 24 through 27 years old. Public Court Records. You're also losing those. So, 2021 suing everyone... August 17, 2026, up-to-date Court Record that YOU brought to YOUR Turf, Bipartisan Judges, Bipartisan Lawyers, a LOSS unable to prove Defamation because you Pled Guilty.... To... September 7 post: "Anyone can say or believe what they want, including truthers and influencers." To... September 17 video: "I have been wrong about dates.... my sources gave me wrong dates"... backpedaling because you're impersonating someone you're not... to... September 19 video: "Defamation / Get my Lawyers"... You're trapped. One. This isn't about your past as you Pled Guilty to a Corruption of Minors as a Plea Bargain due to your text message to the girl that you would commit suicide if she testified against you. You were desperate because you were facing MULTIPLE First Degree Felonies with 20 to 40 years in prison and minimum of 25 years Sex Offender Registration. You waived your Constitutional Rights to a Trial to prove your innocence when you Pled Guilty. This is about you impersonating someone you are not which is a Federal Offense, Title 18 Section 912, Uniform Code of Military Justice Article 106, and the Stolen Valor Act of 2013. Since your impersonation is so vague - if any of your fraudulent transactions links to Human Trafficking - you can be charged with Crimes Against Humanity as Human Trafficking is classified as such. Digital Assets Fraud ties directly to Human Trafficking. You're running Digital Scams with your Tokens, Zoom Meeting Echo Chambers, Zoom Training, etc. Those are outlined as the VERY definitions of Affinity Fraud and "Pig Butchering" found in all Laws, Orders, and National Emergencies via the FBI, SEC, FinCEN, and Investor Government Alerts. Two. You cannot sue anyone moving forward. If you do. Here's what Defense will bring to Discovery: Your "Degrees" in which you do NOT possess shown on your "resume"... Also found in recent Court Records as you were asked about the Master of the Arts and Harvard in which under Oath with REAL Lawyers - you are on record stating they are not real and you have no degree. You are also on Court Record stating you do not know what 'college credit totals' are.

Derek Johnson

19,497 Aufrufe • vor 13 Tagen

HOLY SH*T CLAUDE FILLED A LIMIT AT 29,088.00, EXACT PRICE, EXACT SECOND. CANCELLED THE STOP 0.4 SECONDS LATER. PLACED THE NEXT LIMIT 3 TICKS ABOVE. GOT FILLED AGAIN. +$4,225 STACKED BEFORE THE CANDLE CLOSED. not suggesting trades. not highlighting zones. placing actual orders. filling them. cancelling the ones that don't hit. and moving to the next setup here's what's on screen right now: left side — NQ futures (NQM26) on two timeframes. 1-minute and 30-minute charts running side by side. price at 29,238. multiple limit orders stacked across the chart: → green limit at +$1,869.00 — active → green limit at +$4,225.00 — active → green limit at +$8,335.00 — active → red stop at -$8,765.00 — risk defined → red stop at -$5,785.00 — tighter risk on a separate entry every order has a calculated target and a calculated stop. nothing is open-ended. nothing is "let it run and hope" bottom left — order confirmations updating live: "Order Filled: +1 MNQ Limit at 29,088.00" "Order Cancelled: +1 MNQ Stop Market at 29,075.75" Claude filled the limit. saw the stop was no longer needed. cancelled it. moved on top right — Gamma Exposure heatmap by strike. key gamma strike highlighted at 29,000. the chart shows where dealer positioning creates support and resistance — not where a line was drawn, where actual options hedging forces the market to react bottom right — MotiveWave footprint on MESM6 Range(20) with DOM. bid/ask delta visible at every price level. volume profile stacking on the right. the footprint confirms whether the move into Claude's limit was real buying or just air Claude cross-references the gamma strike map with the footprint. enters at levels where dealer hedging aligns with real order flow. sets mechanical targets. sets mechanical stops. fills, cancels, adjusts — all without a human touching the keyboard a properly configured strategy like this can generate up to $2,000 per day starting from as little as $100 — Claude isn't making one big bet, it's stacking precise micro entries at levels where the options market and the order book agree you're watching charts and thinking about entries Claude already placed five, filled two, cancelled one, and is scanning for the next save this — full strategy breakdown in the article below ↓
0:20

HOLY SH*T CLAUDE FILLED A LIMIT AT 29,088.00, EXACT PRICE, EXACT SECOND. CANCELLED THE STOP 0.4 SECONDS LATER. PLACED THE NEXT LIMIT 3 TICKS ABOVE. GOT FILLED AGAIN. +$4,225 STACKED BEFORE THE CANDLE CLOSED. not suggesting trades. not highlighting zones. placing actual orders. filling them. cancelling the ones that don't hit. and moving to the next setup here's what's on screen right now: left side — NQ futures (NQM26) on two timeframes. 1-minute and 30-minute charts running side by side. price at 29,238. multiple limit orders stacked across the chart: → green limit at +$1,869.00 — active → green limit at +$4,225.00 — active → green limit at +$8,335.00 — active → red stop at -$8,765.00 — risk defined → red stop at -$5,785.00 — tighter risk on a separate entry every order has a calculated target and a calculated stop. nothing is open-ended. nothing is "let it run and hope" bottom left — order confirmations updating live: "Order Filled: +1 MNQ Limit at 29,088.00" "Order Cancelled: +1 MNQ Stop Market at 29,075.75" Claude filled the limit. saw the stop was no longer needed. cancelled it. moved on top right — Gamma Exposure heatmap by strike. key gamma strike highlighted at 29,000. the chart shows where dealer positioning creates support and resistance — not where a line was drawn, where actual options hedging forces the market to react bottom right — MotiveWave footprint on MESM6 Range(20) with DOM. bid/ask delta visible at every price level. volume profile stacking on the right. the footprint confirms whether the move into Claude's limit was real buying or just air Claude cross-references the gamma strike map with the footprint. enters at levels where dealer hedging aligns with real order flow. sets mechanical targets. sets mechanical stops. fills, cancels, adjusts — all without a human touching the keyboard a properly configured strategy like this can generate up to $2,000 per day starting from as little as $100 — Claude isn't making one big bet, it's stacking precise micro entries at levels where the options market and the order book agree you're watching charts and thinking about entries Claude already placed five, filled two, cancelled one, and is scanning for the next save this — full strategy breakdown in the article below ↓

INSIDER

27,518 Aufrufe • vor 22 Tagen

Most social media tools suck. (I've tried them all over 7 years of doing this). They claim to help you go viral. They claim that you'll grow fast on social media. They claim that they'll make it easy on you, so you can just grow your following without having any form of skill or understanding. And now they're trying to go fully agentic, which only works if you know what you're doing (we'll talk about that later, we built the MCP for social media). Plus, when everyone has an advantage, it is no longer an advantage. I promise you, I love AI, but having it write everything for you, especially as a beginner, is a death sentence. You may get lucky and have one, two, or three posts do well. You may gain a lot of followers (because that's what they promise to get you to buy), but then what? Can you replicate it? Does your audience actually care about who you are? Or are you just a one hit wonder like every other creator who tries to get-followers-quick? The creators you love... the ones who put out ideas that actually change your life, understand that the creative process (capturing, resurfacing, and synthesizing ideas) is your competitive edge when everyone is racing to put out as much slop as possible. So we built an actual solution. For the people who know deep down that the toys and gimmicks being marketed to you aren't a long term strategy. It's called Eden. The second brain for creators. It has every scheduling, analytics, auto-dm, and outlier research (for all platforms) features that everyone else has - built by someone who knows what's useful and what's not. But it also has the ideation process built in, which cannot be skipped. Here's what it looks like, and how to write a post people actually care about: 1) Save inspiration from anywhere (and read + highlight) Capturing ideas, posts, links, and highlights inside of an agent or AI chat sounds like hell. Not everything needs to, or should be, jammed into a chat UI. In Eden you can: - Paste any social post link (we transcribe any post) - Integrate with Kindle, Snipd, or Readwise - Pull in X or Substack bookmarks - Jot down quick ideas or write inside docs - Open any YouTube transcript or article in Reader mode and highlight it We auto-categorize and tag everything for you, so our Deep Synthesis agent can find novel connections and relations across your library - making it easy to surface idaes you thought you lost. 2) Research what's working, study top creators Outlier tools are the bread and butter for every creator. When you do research for a YouTube video, you look for high performing titles and thumbnails, and you study what worked for others. But what about every other platform? We have a social corpus of 3m+ posts across all platforms and growing. You can scroll the Discover page for inspiration, search any creator and filter by top posts, or use Deep Social in the AI chat to search across all of them. This is what gives the MCP access to trending topics without scraping millions of posts yourself. 3) Create with everything in one place, build swipe files I don't think I will ever give up writing. I love sitting down in the morning, opening up a page, researching good ideas, and turning it into something worth sharing. In Eden, you work on Boards. You can paste social posts and build shareable swipe files, or you can add your outlines and drafts while having your library, chat, or ideas open right next to it. 4) Chat with anything to get the information you need Claude and ChatGPT are essential, but they are blind. They can't read social links. They can't see what's trending. When you're writing scripts, articles, posts, or drafts - you need to be able to pull information from the sources you research. In today's world, that usually means social posts, YouTube videos, and articles. 5) Draft, schedule, and publish to all platforms X, Instagram, LinkedIn, Substack, Threads, TikTok, and YouTube shorts. And yes, we are one of the first to support article scheduling to X and Substack. You don't have to copy paste to every platform anymore. You simply write, schedule, and let it go out to all platforms. 6) Or do it all from the MCP I don't believe that the future of content is fully agentic. But, that doesn't mean that AI can't help with researching top posts, ideation, and scheduling content. You do the writing and thinking, let AI handle the busy work and resurfacing of ideas you need to find. With Eden, you don't need to give up your Claude, Hermes, or other agentic workflows. You simply connect the MCP and now you can build your own agentic content system, if that's your cup of tea. Eden is the Social Media MCP. You don't have to pay for scrapers anymore. You can finally access trends, any social link, and any account. You can access your analytics, build content dashboards with our infrastructure (that you don't need to maintain as a full time job), and post. We've actually built 4 skills for deep topic research, personal brand strategy, building a content command center, and more - we'll link those below. All in all: If you're a creator, Eden is your base. It's where you write, research, and publish every single day. Try it free below:
1:00

Most social media tools suck. (I've tried them all over 7 years of doing this). They claim to help you go viral. They claim that you'll grow fast on social media. They claim that they'll make it easy on you, so you can just grow your following without having any form of skill or understanding. And now they're trying to go fully agentic, which only works if you know what you're doing (we'll talk about that later, we built the MCP for social media). Plus, when everyone has an advantage, it is no longer an advantage. I promise you, I love AI, but having it write everything for you, especially as a beginner, is a death sentence. You may get lucky and have one, two, or three posts do well. You may gain a lot of followers (because that's what they promise to get you to buy), but then what? Can you replicate it? Does your audience actually care about who you are? Or are you just a one hit wonder like every other creator who tries to get-followers-quick? The creators you love... the ones who put out ideas that actually change your life, understand that the creative process (capturing, resurfacing, and synthesizing ideas) is your competitive edge when everyone is racing to put out as much slop as possible. So we built an actual solution. For the people who know deep down that the toys and gimmicks being marketed to you aren't a long term strategy. It's called Eden. The second brain for creators. It has every scheduling, analytics, auto-dm, and outlier research (for all platforms) features that everyone else has - built by someone who knows what's useful and what's not. But it also has the ideation process built in, which cannot be skipped. Here's what it looks like, and how to write a post people actually care about: 1) Save inspiration from anywhere (and read + highlight) Capturing ideas, posts, links, and highlights inside of an agent or AI chat sounds like hell. Not everything needs to, or should be, jammed into a chat UI. In Eden you can: - Paste any social post link (we transcribe any post) - Integrate with Kindle, Snipd, or Readwise - Pull in X or Substack bookmarks - Jot down quick ideas or write inside docs - Open any YouTube transcript or article in Reader mode and highlight it We auto-categorize and tag everything for you, so our Deep Synthesis agent can find novel connections and relations across your library - making it easy to surface idaes you thought you lost. 2) Research what's working, study top creators Outlier tools are the bread and butter for every creator. When you do research for a YouTube video, you look for high performing titles and thumbnails, and you study what worked for others. But what about every other platform? We have a social corpus of 3m+ posts across all platforms and growing. You can scroll the Discover page for inspiration, search any creator and filter by top posts, or use Deep Social in the AI chat to search across all of them. This is what gives the MCP access to trending topics without scraping millions of posts yourself. 3) Create with everything in one place, build swipe files I don't think I will ever give up writing. I love sitting down in the morning, opening up a page, researching good ideas, and turning it into something worth sharing. In Eden, you work on Boards. You can paste social posts and build shareable swipe files, or you can add your outlines and drafts while having your library, chat, or ideas open right next to it. 4) Chat with anything to get the information you need Claude and ChatGPT are essential, but they are blind. They can't read social links. They can't see what's trending. When you're writing scripts, articles, posts, or drafts - you need to be able to pull information from the sources you research. In today's world, that usually means social posts, YouTube videos, and articles. 5) Draft, schedule, and publish to all platforms X, Instagram, LinkedIn, Substack, Threads, TikTok, and YouTube shorts. And yes, we are one of the first to support article scheduling to X and Substack. You don't have to copy paste to every platform anymore. You simply write, schedule, and let it go out to all platforms. 6) Or do it all from the MCP I don't believe that the future of content is fully agentic. But, that doesn't mean that AI can't help with researching top posts, ideation, and scheduling content. You do the writing and thinking, let AI handle the busy work and resurfacing of ideas you need to find. With Eden, you don't need to give up your Claude, Hermes, or other agentic workflows. You simply connect the MCP and now you can build your own agentic content system, if that's your cup of tea. Eden is the Social Media MCP. You don't have to pay for scrapers anymore. You can finally access trends, any social link, and any account. You can access your analytics, build content dashboards with our infrastructure (that you don't need to maintain as a full time job), and post. We've actually built 4 skills for deep topic research, personal brand strategy, building a content command center, and more - we'll link those below. All in all: If you're a creator, Eden is your base. It's where you write, research, and publish every single day. Try it free below:

DAN KOE

133,459 Aufrufe • vor 1 Monat

What's next for OpenTUI? Here's a technical write-up. Over the last few months OpenTUI gained a lot of stability improvements, new unnecessary but fun features like live audio streaming, and useful features like rendering to the scrollback buffer mixed with a live TUI, called footer mode. Overall the feature set enables building large and complex applications. React and Solid make it super simple and convenient. There is still so much to do though. Three big milestones we have set out to achieve are: - Moving most of the behavioural logic currently living in TypeScript down to the native Zig core - Node compatibility - Optimizing the hell out of primitives like text rendering The render tree mechanisms are currently only usable from TypeScript. Think of the DOM, but controllable like a scene graph. Elements in the render tree are called renderables. They can expose a render method to draw themselves. All renderables are derived from a BaseRenderable. Renderables and the render tree will become native primitives. Building blocks usable from any language bindings. Reducing the TypeScript bindings to a very thin layer, with all the behavioural logic living in the native binary. Moving this down is not just a matter of porting TypeScript classes to Zig. TypeScript currently owns the tree, dirty-state propagation, layout reads, culling, and render ordering. If it still has to walk every node and call into native code for each step, we keep most of the complexity and add FFI overhead. Whole passes and their state need to move together. We took a big step towards this recently by building yoga-layout into the native binary. It exposes part of the official yoga-layout TypeScript package via FFI. Only the API surface that is actually used by OpenTUI. Covered by the test suite of the original yoga-layout package. This already gave a median speedup of ~2.5x, and up to 30x for narrow scenarios. The yoga-layout integration is useful beyond the speedup. Built-in text and editor measurement can now happen entirely in native code during layout instead of calling back into JavaScript. I ran an experiment last month taking this even further, having GPT 5.6 port yoga-layout from C++ to Zig, which gave extremely good results. It would be a burden to maintain right now though, so that's off the table for now. I might come back to it. Simon Klee is working relentlessly on Node compatibility and already has a full Node version of OpenCode running. Node got FFI support in v26.4.0, thanks to help from the Node community, namely Matteo Collina and Paolo Insogna. Behaviour and interfaces seem similar between Node and Bun, but there are some major differences. To get the best performance out of the Node FFI implementation, its usage has to follow some rules. Node has three ways to call native functions: the generic C++/libffi path, the SharedBuffer path, and the V8 Fast API. The generic path converts every argument in Node's C++ layer and then calls the function through libffi. It is flexible, but also the slowest option for frequently called functions. The SharedBuffer path is a middle ground. JavaScript writes scalar values and BigInt pointers into a small per-function buffer, reducing some conversion work. The actual native call still goes through libffi though. Typed arrays used as pointers cannot be packed into this buffer and fall back to the generic path. The path we really want is the V8 Fast API. Node generates a small machine-code trampoline for the exact function signature, allowing optimized JavaScript to call the native function without going through the generic converter or libffi. This only applies to JavaScript-to-native calls. Callbacks from native code into JavaScript still use libffi closures. Getting onto this path is quite strict. A signature can have at most eight arguments and everything must fit into CPU registers. x86-64 Unix systems have room for six GP (general-purpose) and eight FP (floating-point) arguments. AArch64 has room for seven GP and eight FP arguments. Anything that spills onto the stack falls back to a slower path. These are Node fast-path restrictions, not general FFI restrictions. Bun also does not support passing structs by value through its current FFI API. OpenTUI uses bun-ffi-structs to pack ABI-aligned struct data into an ArrayBuffer and passes a pointer instead. Despite the name, the package also works with Node. Pointers need some care too. Typed arrays and ArrayBuffers normally have to be resolved into BigInt addresses first. Eligible functions with exactly one pointer argument get another Fast API entrypoint that can extract the address directly from the buffer. An eligible signature is still not enough. V8 has to optimize a direct call with a fixed number of consistently typed arguments. Wrappers that collect arguments and forward them using spread or Reflect.apply can hide that call shape and keep the function on a slower path. The practical rules are: keep hot signatures within register limits, use direct fixed-arity calls with stable argument types, reuse owned buffers safely, and batch small operations. Then measure the real call site, because eligibility only makes a function fast-capable. We have to design the ABI around these constraints where it makes sense and gives the expected performance improvement. The third big area is text rendering. Today a Text renderable accepts a string, StyledText, or a tree of TextNodes. Before rendering, the TextNode tree is walked and flattened into styled chunks. Those chunks are packed in TypeScript, sent through FFI, copied into a native TextBuffer, and stored in a rope. Styles are represented separately as highlights. A TextBufferView then wraps the rope into visual lines, which are drawn into the visible buffer. This works, but updates are much more expensive than they should be. setStyledText effectively throws away and rebuilds the rope, copies and reparses all text and recreates the style highlights. Changing one TextNode also walks and flattens the complete tree before going through this path again. Text and style segments should instead live directly in the rope and support incremental replacement. Memory ownership is split between retained JavaScript buffers, the native memory registry, rope arenas, wrapping caches, styled-text storage, and highlights. Different operations preserve or reset different parts of that state. This is hard to reason about and can retain memory for much longer than expected. Text storage needs clearer ownership, with fewer lifetimes split across JavaScript and native code. The public API reflects the same split. The t template literal is convenient, but creates another intermediate chunk representation that is mutable, not cached, and not merged. Text also maintains both StyledText content and a special TextNode tree, which do not compose properly. TextNode is only a style scope, not a normal layout primitive, so Text renderables cannot naturally compose inside each other. I think this should become one Text primitive backed directly by rope segments. The template literal API might disappear or become a very thin helper around those native segments. Editing has another temporary layer in TypeScript. Extmarks currently monkey-patch editing operations, scan and adjust all marks after changes, maintain their own undo state, and recreate native highlights. They should become native marks anchored directly in the rope. A proper mark tree, similar to Neovim's marktree, could update marks together with edits, undo, and redo, and provide the foundation for highlights and concealment. Text wrapping has also become too complex. Supporting CJK, emoji, combining characters, ZWJ sequences, tabs, and different terminal width rules currently mixes byte offsets, grapheme indexes, and display-cell columns across several custom algorithms. Dirty views rewrap the complete document. Measurement and drawing can repeat some of the same work. The wrapping implementation needs an overhaul, but the exact shape is still open. The goal is to make Unicode handling easier to maintain, avoid repeated full-document work, and clearly separate byte offsets, graphemes, and terminal display cells. None of this will happen as one big rewrite. We will replace pieces when we understand the problem well enough and when the result is clearly simpler, faster, or more useful. To achieve all of this we might break public interfaces. Thanks to OpenCode and a lot of good models, migration to a new version with breaking changes mostly is not an issue anymore. What do you want to see next for OpenTUI?
0:57

What's next for OpenTUI? Here's a technical write-up. Over the last few months OpenTUI gained a lot of stability improvements, new unnecessary but fun features like live audio streaming, and useful features like rendering to the scrollback buffer mixed with a live TUI, called footer mode. Overall the feature set enables building large and complex applications. React and Solid make it super simple and convenient. There is still so much to do though. Three big milestones we have set out to achieve are: - Moving most of the behavioural logic currently living in TypeScript down to the native Zig core - Node compatibility - Optimizing the hell out of primitives like text rendering The render tree mechanisms are currently only usable from TypeScript. Think of the DOM, but controllable like a scene graph. Elements in the render tree are called renderables. They can expose a render method to draw themselves. All renderables are derived from a BaseRenderable. Renderables and the render tree will become native primitives. Building blocks usable from any language bindings. Reducing the TypeScript bindings to a very thin layer, with all the behavioural logic living in the native binary. Moving this down is not just a matter of porting TypeScript classes to Zig. TypeScript currently owns the tree, dirty-state propagation, layout reads, culling, and render ordering. If it still has to walk every node and call into native code for each step, we keep most of the complexity and add FFI overhead. Whole passes and their state need to move together. We took a big step towards this recently by building yoga-layout into the native binary. It exposes part of the official yoga-layout TypeScript package via FFI. Only the API surface that is actually used by OpenTUI. Covered by the test suite of the original yoga-layout package. This already gave a median speedup of ~2.5x, and up to 30x for narrow scenarios. The yoga-layout integration is useful beyond the speedup. Built-in text and editor measurement can now happen entirely in native code during layout instead of calling back into JavaScript. I ran an experiment last month taking this even further, having GPT 5.6 port yoga-layout from C++ to Zig, which gave extremely good results. It would be a burden to maintain right now though, so that's off the table for now. I might come back to it. Simon Klee is working relentlessly on Node compatibility and already has a full Node version of OpenCode running. Node got FFI support in v26.4.0, thanks to help from the Node community, namely Matteo Collina and Paolo Insogna. Behaviour and interfaces seem similar between Node and Bun, but there are some major differences. To get the best performance out of the Node FFI implementation, its usage has to follow some rules. Node has three ways to call native functions: the generic C++/libffi path, the SharedBuffer path, and the V8 Fast API. The generic path converts every argument in Node's C++ layer and then calls the function through libffi. It is flexible, but also the slowest option for frequently called functions. The SharedBuffer path is a middle ground. JavaScript writes scalar values and BigInt pointers into a small per-function buffer, reducing some conversion work. The actual native call still goes through libffi though. Typed arrays used as pointers cannot be packed into this buffer and fall back to the generic path. The path we really want is the V8 Fast API. Node generates a small machine-code trampoline for the exact function signature, allowing optimized JavaScript to call the native function without going through the generic converter or libffi. This only applies to JavaScript-to-native calls. Callbacks from native code into JavaScript still use libffi closures. Getting onto this path is quite strict. A signature can have at most eight arguments and everything must fit into CPU registers. x86-64 Unix systems have room for six GP (general-purpose) and eight FP (floating-point) arguments. AArch64 has room for seven GP and eight FP arguments. Anything that spills onto the stack falls back to a slower path. These are Node fast-path restrictions, not general FFI restrictions. Bun also does not support passing structs by value through its current FFI API. OpenTUI uses bun-ffi-structs to pack ABI-aligned struct data into an ArrayBuffer and passes a pointer instead. Despite the name, the package also works with Node. Pointers need some care too. Typed arrays and ArrayBuffers normally have to be resolved into BigInt addresses first. Eligible functions with exactly one pointer argument get another Fast API entrypoint that can extract the address directly from the buffer. An eligible signature is still not enough. V8 has to optimize a direct call with a fixed number of consistently typed arguments. Wrappers that collect arguments and forward them using spread or Reflect.apply can hide that call shape and keep the function on a slower path. The practical rules are: keep hot signatures within register limits, use direct fixed-arity calls with stable argument types, reuse owned buffers safely, and batch small operations. Then measure the real call site, because eligibility only makes a function fast-capable. We have to design the ABI around these constraints where it makes sense and gives the expected performance improvement. The third big area is text rendering. Today a Text renderable accepts a string, StyledText, or a tree of TextNodes. Before rendering, the TextNode tree is walked and flattened into styled chunks. Those chunks are packed in TypeScript, sent through FFI, copied into a native TextBuffer, and stored in a rope. Styles are represented separately as highlights. A TextBufferView then wraps the rope into visual lines, which are drawn into the visible buffer. This works, but updates are much more expensive than they should be. setStyledText effectively throws away and rebuilds the rope, copies and reparses all text and recreates the style highlights. Changing one TextNode also walks and flattens the complete tree before going through this path again. Text and style segments should instead live directly in the rope and support incremental replacement. Memory ownership is split between retained JavaScript buffers, the native memory registry, rope arenas, wrapping caches, styled-text storage, and highlights. Different operations preserve or reset different parts of that state. This is hard to reason about and can retain memory for much longer than expected. Text storage needs clearer ownership, with fewer lifetimes split across JavaScript and native code. The public API reflects the same split. The t template literal is convenient, but creates another intermediate chunk representation that is mutable, not cached, and not merged. Text also maintains both StyledText content and a special TextNode tree, which do not compose properly. TextNode is only a style scope, not a normal layout primitive, so Text renderables cannot naturally compose inside each other. I think this should become one Text primitive backed directly by rope segments. The template literal API might disappear or become a very thin helper around those native segments. Editing has another temporary layer in TypeScript. Extmarks currently monkey-patch editing operations, scan and adjust all marks after changes, maintain their own undo state, and recreate native highlights. They should become native marks anchored directly in the rope. A proper mark tree, similar to Neovim's marktree, could update marks together with edits, undo, and redo, and provide the foundation for highlights and concealment. Text wrapping has also become too complex. Supporting CJK, emoji, combining characters, ZWJ sequences, tabs, and different terminal width rules currently mixes byte offsets, grapheme indexes, and display-cell columns across several custom algorithms. Dirty views rewrap the complete document. Measurement and drawing can repeat some of the same work. The wrapping implementation needs an overhaul, but the exact shape is still open. The goal is to make Unicode handling easier to maintain, avoid repeated full-document work, and clearly separate byte offsets, graphemes, and terminal display cells. None of this will happen as one big rewrite. We will replace pieces when we understand the problem well enough and when the result is clearly simpler, faster, or more useful. To achieve all of this we might break public interfaces. Thanks to OpenCode and a lot of good models, migration to a new version with breaking changes mostly is not an issue anymore. What do you want to see next for OpenTUI?

kmdr

31,132 Aufrufe • vor 2 Monaten

Tyrant Georgia troopers get their feelings hurt by someone that knows their rights so they start fishing for a crime to retaliate. ​The video kicks off with the driver exercising his rights as he rolls his window down just an inch or two. When the trooper asks him to roll it down, the driver flatly refuses, stating he is "comfortable right there." ​For drivers, this strategy serves a very specific purpose—it limits face-to-face contact, stops officers from claiming they "smelled alcohol or marijuana," and prevents them from reaching inside the cabin. ​The second the troopers realize they aren't getting standard compliance, the narrative completely changes. Initially, they openly admit they are just "out looking for drunks." But when the driver stands his ground, they immediately pivot to a secondary regulatory justification to keep the stop alive, claiming they want him to pull over for a vehicle inspection because his tires looks "almost too slick" and "almost illegal." ​Once the driver is isolated in secondary, the troopers begin a systematic check of the vehicle's exterior equipment. They order the driver to activate his right blinker, left blinker, brake lights, reverse lights, and high beams and etc. ​While the driver complies with the mechanical commands, he holds his boundary where it counts: when invited to step out of the car to look at his tires, he explicitly declines, stating he feels much safer remaining inside his vehicle. ​The interaction takes a wild turn when the troopers realize this isn't the driver's first run-in with their department. One trooper asks him if he remembers Trooper Mike Freeman writing him a ticket for doing 84 mph in a 65 mph zone on Interstate 575. The driver acknowledges the past ticket, keeping his composure entirely. ​Ultimately, because the driver knows his rights, keeps his cool, and his vehicle's equipment actually functions perfectly, the troopers have no legal grounds to push further. They are forced to hand back his license and let him roll. ​What's your take on this encounter? Is this a legitimate safety check, or are we looking at officers who got their feelings hurt by a window crack and resorted to an equipment checklist just to find something to cite him for? This is text book tyrant cops right here in my opinion.
4:51

Tyrant Georgia troopers get their feelings hurt by someone that knows their rights so they start fishing for a crime to retaliate. ​The video kicks off with the driver exercising his rights as he rolls his window down just an inch or two. When the trooper asks him to roll it down, the driver flatly refuses, stating he is "comfortable right there." ​For drivers, this strategy serves a very specific purpose—it limits face-to-face contact, stops officers from claiming they "smelled alcohol or marijuana," and prevents them from reaching inside the cabin. ​The second the troopers realize they aren't getting standard compliance, the narrative completely changes. Initially, they openly admit they are just "out looking for drunks." But when the driver stands his ground, they immediately pivot to a secondary regulatory justification to keep the stop alive, claiming they want him to pull over for a vehicle inspection because his tires looks "almost too slick" and "almost illegal." ​Once the driver is isolated in secondary, the troopers begin a systematic check of the vehicle's exterior equipment. They order the driver to activate his right blinker, left blinker, brake lights, reverse lights, and high beams and etc. ​While the driver complies with the mechanical commands, he holds his boundary where it counts: when invited to step out of the car to look at his tires, he explicitly declines, stating he feels much safer remaining inside his vehicle. ​The interaction takes a wild turn when the troopers realize this isn't the driver's first run-in with their department. One trooper asks him if he remembers Trooper Mike Freeman writing him a ticket for doing 84 mph in a 65 mph zone on Interstate 575. The driver acknowledges the past ticket, keeping his composure entirely. ​Ultimately, because the driver knows his rights, keeps his cool, and his vehicle's equipment actually functions perfectly, the troopers have no legal grounds to push further. They are forced to hand back his license and let him roll. ​What's your take on this encounter? Is this a legitimate safety check, or are we looking at officers who got their feelings hurt by a window crack and resorted to an equipment checklist just to find something to cite him for? This is text book tyrant cops right here in my opinion.

Giggling Gannon

98,197 Aufrufe • vor 3 Monaten