Useful information.

Useful information.

Useful information.

🚨Improve YOUR🫵PC with this Optimization🚨 ✅Disable Telemetry, Advertising, Data... Collection and other unnecessary things hogging your CPU. 📌More useful information in my Discord Server where I talk about how to ACTUALLY Lower Ping and more. ❤️+♻️show more

I have found it really useful to convert notes/essays/paper... into Artifacts. And it seems Fable 5 is really good at it. Artifacts are great for building deeper intuition on any topic. Just shared this one I did on Satya's latest essay on the Reverse Information Paradox.show more

There are a LOT of amazing "pose reference packs"... for artists. I love them. Wide variety of people and a wide variety of costumes etc. Incredibly useful for drawing from. Also, highly recommended to draw from still frames of whatever actiony videos you can find. Poses are great - but there's information in movement too. Here's one that's fun - by Elsaucepapii on Facebook. Set it to half speed. Scroll it. Stop on a frame. Draw.show more

The #CoreIgnitionDrop Season 1 is ending today, September 11th... at 18:29 UTC and Season 2 is starting tomorrow, September 12th at 00:00 UTC🔥 Here is some useful information: 🔸Season 1 rewards will remain claimable (for 6 months as previously announced); 🔸Season 1 badges and invitees will be carried over to Season 2; 🔸All Season 1 participants who don't have an invite code will have one once logged in for Season 2; 🔸Sparks will be reset to 0 when Season 2 starts; 🔸The 3rd claiming phase for Season 1 will take place after Season 2 starts. Stay tuned and get ignited! 🔗show more

Force feedback demo Force feedback is when joystick is... pushing on your hand when something is pushing on the robot arm. Feeling the force - so much helpful to control the robot, that done well it allows you to do tasks even without visual feed. You can make an experiment: close your eyes - you can easily get the headphones out of the case. Also, visual information is often not enough. For example, you're trying to pull out a usb connector, but you pull it at the wrong angle, causing it to get stuck. Visually, nothing changes, but the pressure is intense and you can break the connector. Surgical robots have been using force feedback for years, and there are also 3D styluses which use this feature, proving that the technology works and is useful. But in modern robots with AI, it's hardly ever implemented. Although it's useful for both teleoperation and AI model. That's one of the reasons why we are building our robotic arms starting with off the shelf motors rather than taking the whole off the shelf arm. There are still a range of easy wins that can be made iterating robot hardware.show more

A Nairobi man is seeking help in tracing two... women suspected of drugging and robbing him of valuables worth over Ksh 700,000 after he invited the suspects to his residence following a night out last weekend in Ongata Rongai, Kajiado County. According to available information and CCTV footage, one of the women was seen stepping outside to conduct what appears to be surveillance, while her accomplice allegedly packed the stolen items inside the house. The suspects made away with a silver Apple MacBook Pro 16-inch, an iPhone 14 Pro Max, an iPod Pro Max, a pair of Apple AirPods Max headphones, and KSh 52,000 which was transferred from the victim’s M-PESA account to a phone number that has since been switched off. The footage shows the two women leaving the premises at around 2:55 PM on Saturday while carrying bags believed to contain the stolen items. The victim was reportedly unconscious inside the apartment at the time of the theft. He is now appealing to the public for assistance in identifying the two suspects, hoping that the CCTV footage will prompt anyone with useful information to come forward.show more

LLM Artifacts Connected to Andrej Karpathy's LLM Knowledge base... idea, I've been building out a fun way to generate dynamic artifacts from these knowledge bases with the goal of discovering and revealing meaningful and deeper insights. LLM KBs are hard to consume for humans, as I think they are more built for agents. So the question is, what form would be useful for humans to take actions and make important decisions? That's what I am trying to figure out with these artifacts. The artifact example shows a pulse on HN discussions around AI-related stories. The insights can go deeper, of course, but this is already super fun and thought-provoking, like some of my favorite podcasts. The format and depth matter a lot. The aggregation skills of agents are outstanding if you tune the prompts and skill carefully. I built this artifact generator in a few minutes through an agent skill, but I feel like there are so many ways that LLM-generated information can be used and consumed. Like generating deeper insights and analysis, and things that are just not feasible for humans today. The generated artifact (including its data and design) serves as reusable templates or can be updated in real-time via auomations, which is something I am also working on. It is truly an insane way to monitor and track information. Better than a newsletter. Better than newspapers. There is something about this that gets me really excited about the future of AI agents for knowledge generation and discovery. Lots of hidden gems everywhere just waiting to be discovered and acted on if the information is presented correctly. This is not perfect. The format, style/prose can be improved, but this is easy to customize via skill. You can personalize it to your liking. I feel like these dynamic artifacts are going to emerge as a strong new medium to stay on the cutting edge of things, both for agents and humans. My target is research, of course. This was just a basic example. Besides animation, I am also targeting other components like voice, videos, images, slides, etc. This space is full of opportunities to explore. Skill for this coming soon.show more

AI has had exactly two scaling axes that worked... so far, and the second one is starting to look finite too the first one was pretraining: with scaling parameters and data, we got world knowledge (i.e. ChatGPT had read enough to know things), but it started saturating a while ago the second one was RL, and people had been doing RL the whole time before that: RLHF is RL but it never scaled far because it was trying to control the exact output, which tokens come out, how the text reads, but you can only push that so far before you’re just polishing RLVR dropped that constraint: giving the model a task, then checking whether the final answer is right, and ignoring everything in between -- so the model does whatever it wants in the middle and only the endpoint gets graded, and that’s much closer to actual RL and it’s what bought us planning and reasoning (arguably, tool use sits around 2.5 on this list -- while useful, it's not a different kind of thing) so one axis gave knowledge, the other gave reasoning, and both of them are one model working alone the next axis is how many models you can get working on the same problem, which is a different kind of axis than the previous two we know that multi-agent RL has always been the harder problem: I spent years in that literature and the gap between single-agent and multi-agent is definitely not incremental -- it’s a whole different class of difficulty! which is also why the derivatives are steep at the start, nobody has picked the easy wins yet... and the thing that gates this multi-agent coordination is communication: models can only coordinate as well as they can exchange information, and right now they do that by writing sentences to each other imagine what could we possibly achieve if we properly open that third axis development by letting models to exchange information in their native "language" without loosing any computational data that they produce during inferenceshow more

Grok Bot is not just another chatbot. It is... an always-on AI teammate with its own computer. It can sign in to your tools, work across apps, inboxes and websites, keep working 24/7 and return with finished work. Here are 100 useful tasks you could give Grok Bot: Sales: • Research companies, customers, leads and competitors • Build targeted lists of potential customers • Find and score high-intent sales prospects • Enrich contact details and company records • Draft personalized emails and LinkedIn messages • Turn call transcripts into CRM notes and follow-ups • Prepare detailed account briefs before meetings • Keep CRM records and customer details updated • Flag stalled deals, risks and overdue next steps • Prepare weekly sales reports and pipeline scoreboards Marketing and content: • Research content ideas and trending topics • Prepare marketing campaign briefs • Create and manage campaigns across different tools • Research competitors’ marketing strategies • Build weekly or monthly content calendars • Repurpose content for social media, email and other platforms • Draft social posts, newsletters and blog content • Create multiple versions of advertisements and copy • Monitor brand mentions and prepare responses • Prepare campaign performance reports Inbox and administration: • Monitor your inbox and organize emails • Draft replies in your preferred writing style • Create a daily summary of important messages • Extract tasks, deadlines and action items from emails • Follow up on stalled conversations and forgotten tasks • Prepare meeting agendas and discussion points • Turn meeting notes into tasks and follow-ups • Organize email attachments and cloud files • Enter information into forms and internal systems • Update project trackers, documents and dashboards Finance and operations: • Process invoices received through Gmail • Extract invoice information automatically • Match invoices with purchase orders • Track expenses and prepare finance updates • Organize receipts and expense claims • Clean and reconcile financial spreadsheets • Compare vendor prices and prepare recommendations • Monitor software subscriptions and renewals • Track inventory and flag low supplies • Handle repetitive office operations Hiring and HR: • Help onboard new employees • Organize seating, accounts and equipment for new hires • Draft job descriptions • Research and identify potential candidates • Organize applications and candidate information • Coordinate interview schedules • Prepare candidate briefs before interviews • Combine interview feedback into one report • Create training materials and onboarding guides • Prepare employee offboarding checklists Customer support and success: • Monitor incoming customer requests • Create and categorize support tickets • Prioritize urgent customer issues • Draft customer support replies • Search internal documentation for answers • Escalate important problems to the correct team • Send product updates to customers • Keep follow-up tasks synchronized across tools • Prepare customer health and renewal reports • Turn customer feedback into clear themes Product and engineering: • Prepare and test product demo environments • Fix stale demo data and create readiness checklists • Reproduce software bugs inside the product interface • File detailed bug reports with steps and evidence • Hand bugs to another debugging Bot to work on fixes • Test websites and apps for common problems • Run repetitive quality-assurance checks • Test logins, forms, buttons and checkout flows • Update project boards and release notes • Turn user feedback into organized feature requests Research and reporting: • Conduct detailed web research • Collect sources and prepare cited summaries • Compare competitors’ products and pricing • Monitor industry news and important announcements • Organize research findings into tables • Clean and analyze large spreadsheets • Create daily, weekly or monthly reports • Update internal dashboards with new information • Prepare presentation outlines and meeting briefs • Summarize long documents, emails and transcripts E-commerce and online business: • Update product listings across platforms • Check inventory and identify low-stock products • Monitor competitors’ prices • Analyze customer reviews and common complaints • Track delayed or problematic orders • Prepare return and refund cases for approval • Follow up with vendors about missing orders • Create promotional campaign drafts • Clean product catalogs and remove outdated information • Prepare daily store performance summaries Events and project management: • Research venues, vendors and service providers • Compare quotes and prepare options • Create event checklists and timelines • Coordinate invitations and attendee lists • Track project deadlines and responsibilities • Remind team members about overdue work • Collect progress updates from different tools • Prepare weekly project status reports • Identify blockers before they delay a project • Keep everyone aligned on the latest information Bots, routines and automation: • Watch you complete a workflow and learn how you do it • Save workflows as reusable routines • Repeat multi-step tasks without being taught again • Learn your voice, preferences and edge cases • Manage several specialist Bots simultaneously • Let Bots communicate and share information • Place Bots in group chats to coordinate work • Use one chief-of-staff Bot to manage the others • Run research, reporting and other jobs overnight • Ask for your approval only when a decision requires you You can message Grok Bot like a coworker, teach it how you prefer things done and let it improve over time. The real value is not just getting an answer. It is handing off an entire job and getting finished work back.show more

My Prediction Market Tools list + Trading Workflow A)... Tools I use: - Betmoar: Filtering + Analysing specific smart money wallets - : Spotting volume and OI changes + whale flows - Polymarket Analytics: Comparing markets across different Prediction Markets - Polycule: Trading on the go - Polysights: Advanced Metrics related to volatility and trends B) My workflow: 1) I usually trade sports + crypto price prediction markets since I have some kind of expertise there 2) I use Betmoar to filter and find interesting markets. Try to enter new markets as early as possible since there are significant mispricings 3) I then use Hashdive to see what OI changes and whale flows are like for a particular market 4) Polymarket Analytics to compare markets across PM and Kalshi to see where I can get the best price 5) Trade directly through PM and Kalshi for now to execute if im on my laptop and Polycule when I'm not 6) I also check in daily on smart money wallets I follow on BetMoar to see if they've made any interesting trades recently I don't focus on Arbitrage, LPing or delta-neutral trades at all. Prefer to either take directional trades where I have an information edge or I'm early Sometimes I'll 'gamble' on sports markets because it's fun C) Takeaways for you: - Only trade in markets you have in-depth knowledge/edge - Being early is lucrative in prediction markets - Pro tools can save you hours trying to find alpha - Don't use too many tools. Use a few and master them instead - Everyone has their own process. Don't copy mine it won't work for you. The point of this tweet was to educate you on how you can build your own process LMK if there's any other useful tools I should check out that could optimise my workflow further!show more

Model-Free Reinforcement Learning (MFRL) has been alluring, especially with... supercharged compute with physics on GPU. However, the methods use 0-th order gradients, and are often not the best optimizers. Can we do better than PPO in continuous control for robotics? Turns out yes! 🥳 tl;dr: Faster, better RL than PPO in continuous control 💪 The answer lies in using more information from the simulation. We are juicing the simulation on GPU as it is, why not use it for gradients as well? This has been a driving question in a series of our works. We first studied this problem in ICLR 2022 paper on Short Horizon Actor Critic Naive gradient based methods are stuck in local minima and have exploding/vanishing gradients. SHAC solved this problem truncated rollouts and model based value estimation, where the model is Differentiable Sim. This boosted sample efficiency and wall-clock time immensely especially in high dimensional systems such as humanoids Yet, given enough compute PPO often caught up. Our follow up paper on on Adaptive Horizon Actor Critic at ICML 2024 discovers the cause and provides a fix. However, we find that even when given ground-truth dynamics, not all gradients are useful due to sample error. 1st-Order Model-Based Reinforcement Learning methods employing differentiable simulation provide gradients with reduced variance but are susceptible to bias in scenarios involving stiff dynamics, such as physical contact. We find that back-propagating through contact and long trajectories drastically reduces gradient accuracy. Using this insight, we propose AHAC to dynamically adapt its roll-out horizon to avoid differentiating through stiff contact. AHAC is a first-order model-based RL algorithm that learns high-dimensional tasks in minutes (wall clock) and outperforms PPO by 40%, even in the limit of data provided to PPO. This work is led by Ignat Georgiev alongside Krishnan Srinivasan, Jie Xu, Eric Heiden and ample assistance from warp team at NVIDIA Robotics (Miles Macklin)show more