Загрузка видео...

Не удалось загрузить видео

Возникла проблема при загрузке этого видео. Это может быть связано с временными проблемами сети или видео может быть недоступно.

На главную

Announcing... CHAMBER TRIALS x BRINKZ DRAFT INVITATIONAL 8 DIFFERENT CAPTAINS DRAFT UP TEAMS OF 4 FOR AWESOME MINECRAFT GAMEPLAY. TAKING PLACE ON FRIDAY (yes friday) JULY 31ST @ 2 PM EST !!! TEAMS SOMETIME NEXT WEEK !!

Brinkz

1,342 subscribers

11,677 просмотров • 1 месяц назад •via X (Twitter)

Anya Rossi• Live Now

Private livecam show

Комментарии: 0

Нет доступных комментариев

Здесь появятся комментарии из оригинального поста

Похожие видео

The LIV Golf League has made major changes to the format of the Team Championship, including a preliminary round to be played on Wednesday between teams ranked 12 and 13 in the season long standings. There will also be no byes for top seeded teams. The losing team from the preliminary round will be fully eliminated and will take no further part in the competition. On Friday, there will be 6 quarter finals matches, where all 12 remaining teams will face off in head to head match play. The captains of the top seeded teams have the opportunity to choose their opponents as in previous years. The format is also the same, with 2 singles matches and 1 foursomes match, with 2 points needed to advance. On Saturday, all 12 teams will be divided into 2 brackets. The Championship Bracket, with the winning teams from Friday, and the Rankings Bracket, with the 6 teams who lost. Once again, seeded team captains can select who they play against and the format remains 2 singles matches and 1 foursomes. On Sunday, a champion will be decided and they revert back to the standard team format of the regular season, individual strokeplay with all 4 scores counting towards the team total. The 3 winning teams from the Championship Bracket will face off for the championship rings and the remaining 9 teams will play in their Rankings Brackets for minor places. Ripper GC, captained by Cam Smith will be looking to defend the title they won last season. The previous winners are Bryson DeChambeau’s Crushers GC and Dustin Johnson’s 4 Aces GC. The 2025 Team Championship will be held at The Cardinal at St John’s Resort in Michigan on August 20th-24th.

The LIV Golf League has made major changes to the format of the Team Championship, including a preliminary round to be played on Wednesday between teams ranked 12 and 13 in the season long standings. There will also be no byes for top seeded teams. The losing team from the preliminary round will be fully eliminated and will take no further part in the competition. On Friday, there will be 6 quarter finals matches, where all 12 remaining teams will face off in head to head match play. The captains of the top seeded teams have the opportunity to choose their opponents as in previous years. The format is also the same, with 2 singles matches and 1 foursomes match, with 2 points needed to advance. On Saturday, all 12 teams will be divided into 2 brackets. The Championship Bracket, with the winning teams from Friday, and the Rankings Bracket, with the 6 teams who lost. Once again, seeded team captains can select who they play against and the format remains 2 singles matches and 1 foursomes. On Sunday, a champion will be decided and they revert back to the standard team format of the regular season, individual strokeplay with all 4 scores counting towards the team total. The 3 winning teams from the Championship Bracket will face off for the championship rings and the remaining 9 teams will play in their Rankings Brackets for minor places. Ripper GC, captained by Cam Smith will be looking to defend the title they won last season. The previous winners are Bryson DeChambeau’s Crushers GC and Dustin Johnson’s 4 Aces GC. The 2025 Team Championship will be held at The Cardinal at St John’s Resort in Michigan on August 20th-24th.

Flushing It

133,357 просмотров • 1 год назад

Matt Ponatoski will soon have to decide whether he’s heading to the MLB or will suit up for Kentucky football AND baseball. 🏈⚾ “I’m going to be fully prepared to go to school and go be at Kentucky and enroll and go compete for a quarterback battle when I get there. But at the same time, like going and putting my best foot forward to these MLB teams and showing them like I have what it takes for you to draft me.” Ponatoski on where he is at in the decision-making. “It’s one A and one B,” Ponatoski said. “I’m going to go have a great career at Kentucky, or I’m going to go be a professional baseball player. It’s like you can’t draw it up much better.” The MLB Draft will be taking place July 11-13 in Philadelphia and Ponatoski will soon have to make a decision in a few months. 😼 #BBN

Matt Ponatoski will soon have to decide whether he’s heading to the MLB or will suit up for Kentucky football AND baseball. 🏈⚾ “I’m going to be fully prepared to go to school and go be at Kentucky and enroll and go compete for a quarterback battle when I get there. But at the same time, like going and putting my best foot forward to these MLB teams and showing them like I have what it takes for you to draft me.” Ponatoski on where he is at in the decision-making. “It’s one A and one B,” Ponatoski said. “I’m going to go have a great career at Kentucky, or I’m going to go be a professional baseball player. It’s like you can’t draw it up much better.” The MLB Draft will be taking place July 11-13 in Philadelphia and Ponatoski will soon have to make a decision in a few months. 😼 #BBN

Chris Beasmore

84,392 просмотров • 4 месяцев назад

Nick Dumesnil (California Baptist Baseball) has as much upside as anyone in the 2025 draft. With question marks early on about how he'd fare against the top arms in the country on the Cape, "Doomey" proved quickly as one of the best players in the summer league circuit. His hands start high with the bat coiled behind his head. While a somewhat unorthodox approach, he's able to get his bat head up and hands lowered upon loading. It's an explosive operation with elite bat speed, averaging as much as 79.9 mph in the CCBL (per @Blast_Bsbl). It's a consistent swing making hard contact and driving the ball to both gaps (led the CCBL with 12 2B). The quick hip hinge allows for a powerful on-time rotation through the ball as he barrels the ball at a high clip (up to 30.4% this summer per Joe Doyle). Then there's the speed. Doomey set the Brewster Whitecap SB record with 26 (27 ATT) in 36 G. His 96.2% success rate is the highest in league history for someone with 20+ steals in a season (per League Statistician). He gets good jumps on the base paths and good reads off the bat defensively. There is some swing and miss to his game. He walks at an average rate (8.6 BB% in the CCBL) and struck out just 12.9% of the time in the WAC but that went up to 21.7% this summer. There's some chase low out of the zone vs break and did way more damage vs LHP (.435/.469/.783; .348 ISO) than RHP (.247/.333/.337; .090 ISO). A true five-tool talent, Nick Dumesnil made a name for himself this summer on the Cape, and will soon be at the top of many teams' draft boards next July.

Nick Dumesnil (California Baptist Baseball) has as much upside as anyone in the 2025 draft. With question marks early on about how he'd fare against the top arms in the country on the Cape, "Doomey" proved quickly as one of the best players in the summer league circuit. His hands start high with the bat coiled behind his head. While a somewhat unorthodox approach, he's able to get his bat head up and hands lowered upon loading. It's an explosive operation with elite bat speed, averaging as much as 79.9 mph in the CCBL (per @Blast_Bsbl). It's a consistent swing making hard contact and driving the ball to both gaps (led the CCBL with 12 2B). The quick hip hinge allows for a powerful on-time rotation through the ball as he barrels the ball at a high clip (up to 30.4% this summer per Joe Doyle). Then there's the speed. Doomey set the Brewster Whitecap SB record with 26 (27 ATT) in 36 G. His 96.2% success rate is the highest in league history for someone with 20+ steals in a season (per League Statistician). He gets good jumps on the base paths and good reads off the bat defensively. There is some swing and miss to his game. He walks at an average rate (8.6 BB% in the CCBL) and struck out just 12.9% of the time in the WAC but that went up to 21.7% this summer. There's some chase low out of the zone vs break and did way more damage vs LHP (.435/.469/.783; .348 ISO) than RHP (.247/.333/.337; .090 ISO). A true five-tool talent, Nick Dumesnil made a name for himself this summer on the Cape, and will soon be at the top of many teams' draft boards next July.

Ethan Kagno

16,658 просмотров • 2 лет назад

my 8 GB VRAM gaming laptop is absolutely going to hate me for this. but I still did it. ran a 31b dense model (Gemma 4 31b Q4) with only 8 GB VRAM last week I ran Gemma 4 26B A4B a mixture of experts model on my RTX 4060 and hit 25–28 tokens/sec using llama.cpp's new MTP support. smooth. snappy. but MoE has a secret: it only activates 4B parameters per token despite having 26B total. that's why it flies. so the real question started haunting me. what if I throw a full, no tricks, every parameter fires on every token, 31B DENSE model at the same machine? # Hardware: GPU: NVIDIA RTX 4060, 8 GB VRAM RAM: 16 GB CPU: Intel Core i7 H Laptop. Gaming. Modest. The model: gemma-4-31B-it-qat-UD-Q4_K_XL.gguf (model's unsloth huggingface link in the comments) This is Google DeepMind's flagship dense model in the Gemma 4 family that can run on single consumer GPU. It packs a hybrid attention architecture, supports up to 256K context natively, and is QAT (Quantization Aware Training) optimized, meaning it retains far more quality than standard post training quants at the same bit depth. This is NOT the MoE. This is 31 BILLION dense parameters, every single one of them loaded. # the flags I used: -m gemma-4-31B-it-qat-UD-Q4_K_XL.gguf -cnv --spec-type draft-mtp --spec-draft-model mtp-gemma-4-31B-it.gguf --spec-draft-n-max 8 --spec-draft-p-min 0.6 -c 6000 -v Multi Token Prediction (MTP) is still active here. Separate draft GGUF required, same as the 26B setup. # Results: → Decode: ~3 tokens/sec → Prefill: ~2 tokens/sec → Context: 6000 tokens → Hardware crying quietly in the corner: yes so is 3 tps actually usable? For real time back and forth chat? Not ideal. You're not having a fluid conversation at 3 tps. but slow ≠ useless. And this is where it gets genuinely interesting. think about how senior devs actually work in a real team. But when something is architectural, deeply complex, or needs serious reasoning? they walk down the hall and escalate to the senior. That's exactly the local AI agent architecture this unlocks: → Fast orchestrator model (Gemma 4 26B MoE at 25+ tps) handles routing, simple queries, tool calls, memory. The junior dev. → Gemma 4 31B dense is the senior, called only when the fast model genuinely hits a wall. Hard multi step reasoning. Complex code generation. Deep architectural decisions. The agentic loop stays fast. Only the hard hops touch the 31B. That's a legitimate production grade local AI architecture on a budget hardware. (requires 2 8gb gpus) other workflows where 3 tps is completely fine: - overnight batch jobs. summarize documents, extract structured data, review code. Fire it off. Sleep. wake up to results. - One shot deep reasoning - Silent code audit loops, you write and test, the 31B reviews diffs and flags issues in the background between your sprints - Any workflow where output quality > output speed A few weeks ago, nobody was running a 30B+ dense model on a single consumer GPU with 8 GB VRAM. At all. Now we're doing it on an Intel i7-H gaming laptop with a NVIDIA RTX 4060, thanks to llama.cpp + QAT quants + MTP speculative drafting. Google DeepMind said the Gemma 4 31B targets "consumer GPUs and workstations." They were not exaggerating. The hardware bar to run serious frontier class models locally keeps dropping. the tools are here. the models are here. you just have to be willing to abuse your laptop a little. what workflows would you actually run on a local 3 tps 31B dense model? genuinely curious. drop it below.

my 8 GB VRAM gaming laptop is absolutely going to hate me for this. but I still did it. ran a 31b dense model (Gemma 4 31b Q4) with only 8 GB VRAM last week I ran Gemma 4 26B A4B a mixture of experts model on my RTX 4060 and hit 25–28 tokens/sec using llama.cpp's new MTP support. smooth. snappy. but MoE has a secret: it only activates 4B parameters per token despite having 26B total. that's why it flies. so the real question started haunting me. what if I throw a full, no tricks, every parameter fires on every token, 31B DENSE model at the same machine? # Hardware: GPU: NVIDIA RTX 4060, 8 GB VRAM RAM: 16 GB CPU: Intel Core i7 H Laptop. Gaming. Modest. The model: gemma-4-31B-it-qat-UD-Q4_K_XL.gguf (model's unsloth huggingface link in the comments) This is Google DeepMind's flagship dense model in the Gemma 4 family that can run on single consumer GPU. It packs a hybrid attention architecture, supports up to 256K context natively, and is QAT (Quantization Aware Training) optimized, meaning it retains far more quality than standard post training quants at the same bit depth. This is NOT the MoE. This is 31 BILLION dense parameters, every single one of them loaded. # the flags I used: -m gemma-4-31B-it-qat-UD-Q4_K_XL.gguf -cnv --spec-type draft-mtp --spec-draft-model mtp-gemma-4-31B-it.gguf --spec-draft-n-max 8 --spec-draft-p-min 0.6 -c 6000 -v Multi Token Prediction (MTP) is still active here. Separate draft GGUF required, same as the 26B setup. # Results: → Decode: ~3 tokens/sec → Prefill: ~2 tokens/sec → Context: 6000 tokens → Hardware crying quietly in the corner: yes so is 3 tps actually usable? For real time back and forth chat? Not ideal. You're not having a fluid conversation at 3 tps. but slow ≠ useless. And this is where it gets genuinely interesting. think about how senior devs actually work in a real team. But when something is architectural, deeply complex, or needs serious reasoning? they walk down the hall and escalate to the senior. That's exactly the local AI agent architecture this unlocks: → Fast orchestrator model (Gemma 4 26B MoE at 25+ tps) handles routing, simple queries, tool calls, memory. The junior dev. → Gemma 4 31B dense is the senior, called only when the fast model genuinely hits a wall. Hard multi step reasoning. Complex code generation. Deep architectural decisions. The agentic loop stays fast. Only the hard hops touch the 31B. That's a legitimate production grade local AI architecture on a budget hardware. (requires 2 8gb gpus) other workflows where 3 tps is completely fine: - overnight batch jobs. summarize documents, extract structured data, review code. Fire it off. Sleep. wake up to results. - One shot deep reasoning - Silent code audit loops, you write and test, the 31B reviews diffs and flags issues in the background between your sprints - Any workflow where output quality > output speed A few weeks ago, nobody was running a 30B+ dense model on a single consumer GPU with 8 GB VRAM. At all. Now we're doing it on an Intel i7-H gaming laptop with a NVIDIA RTX 4060, thanks to llama.cpp + QAT quants + MTP speculative drafting. Google DeepMind said the Gemma 4 31B targets "consumer GPUs and workstations." They were not exaggerating. The hardware bar to run serious frontier class models locally keeps dropping. the tools are here. the models are here. you just have to be willing to abuse your laptop a little. what workflows would you actually run on a local 3 tps 31B dense model? genuinely curious. drop it below.

Alok

63,689 просмотров • 2 месяцев назад

August recap of my Fogochain journey: Faucet →DeFi→Trading→Meme Token 1/ Testnet Expedition. Ep 1–5: ✦ Testnet Expedition Episode #1 : ✦ Testnet Expedition Episode #2 : ✦ Testnet Expedition Episode #3 : ✦ Testnet Expedition Episode #4 : ✦ Testnet Expedition Episode #5 : 2/ Faucet Guides: ✦ How To Get Faucet (Unofficial) : ✦ How to Get Faucet For Free : ✦ How to Claim Faucet by GasZip : 3/ DeFi ✦ Pyron Review : ✦ Trading Experience on Valiant : 4/ Meme Token ✦ Introducing $FF Token : Achievements in August. ▪️Earned Ember & Flame Crafter role on Fogo discord in 1 month ▪️Reached Top 8 Token Holder on Ember Dispatch ▪️Big impressions on X, even from a small account This month, my learn progress wasn’t about a single lucky break. It came from showing up daily, taking notes, fixing mistakes, and sharing what I found so others could move faster, too. Start small today. Return tomorrow. Repeat. Now, I’d love to hear from you: Which content did you enjoy the most, and what should I create next? More tutorials, trading notes, safety checklists, or deep dives into burner ecosystem? Your ideas will shape the next chapter of this journey.

August recap of my Fogochain journey: Faucet →DeFi→Trading→Meme Token 1/ Testnet Expedition. Ep 1–5: ✦ Testnet Expedition Episode #1 : ✦ Testnet Expedition Episode #2 : ✦ Testnet Expedition Episode #3 : ✦ Testnet Expedition Episode #4 : ✦ Testnet Expedition Episode #5 : 2/ Faucet Guides: ✦ How To Get Faucet (Unofficial) : ✦ How to Get Faucet For Free : ✦ How to Claim Faucet by GasZip : 3/ DeFi ✦ Pyron Review : ✦ Trading Experience on Valiant : 4/ Meme Token ✦ Introducing $FF Token : Achievements in August. ▪️Earned Ember & Flame Crafter role on Fogo discord in 1 month ▪️Reached Top 8 Token Holder on Ember Dispatch ▪️Big impressions on X, even from a small account This month, my learn progress wasn’t about a single lucky break. It came from showing up daily, taking notes, fixing mistakes, and sharing what I found so others could move faster, too. Start small today. Return tomorrow. Repeat. Now, I’d love to hear from you: Which content did you enjoy the most, and what should I create next? More tutorials, trading notes, safety checklists, or deep dives into burner ecosystem? Your ideas will shape the next chapter of this journey.

Raniii 🔥

65,755 просмотров • 1 год назад

On a conference call to promote American Century Championship golf tournament, Kyle Lowry was asked if he saw any Villanova traits from ‘Nova Knicks in Game 4. Lowry: “(When) Josh Hart missed that layup & Wemby got fouled, if you looked at Jalen Brunson, he went straight to Josh & he said, 'Don’t worry about it, we’ll be OK.' "And I think that’s the one thing that us, Villanova (players)…we all understand the grit and the grind but we all understand (the importance of) having a great positive attitude and that’s what (head coach) Jay (Wright) taught us; have a good attitude.... no matter what (happens)… you’ll figure it out out. (Hart) misses a layup, Wemby gets fouled. Jalen was right there to pick him up and he said, ‘Don’t worry about it, we’ll be arlight.' And ya’ll seen what happened (the rest of the game). That’s just the type of attitude, the type of resilience that Nova brings…I just think (the Nova Knicks) just brought a different type of process and winnng culture for the Knicks." Lowry finished 38th in last year's American Century Championship, which is set for July 8-12 with coverage on NBC/Peacock, Friday–Sunday, July 10-12. He’s looking forward to this year’s tournament. When asked about his plans for next NBA season, Lowry said he doesn't know yet what he’ll do. "We’ll see. I like playing golf so I don’t know what’s next....I've gotta get my game right so I don’t know what’s going to happen (but) I need better odds (than his current odds for the American Century)," Lowry said, lightheartedly.

On a conference call to promote American Century Championship golf tournament, Kyle Lowry was asked if he saw any Villanova traits from ‘Nova Knicks in Game 4. Lowry: “(When) Josh Hart missed that layup & Wemby got fouled, if you looked at Jalen Brunson, he went straight to Josh & he said, 'Don’t worry about it, we’ll be OK.' "And I think that’s the one thing that us, Villanova (players)…we all understand the grit and the grind but we all understand (the importance of) having a great positive attitude and that’s what (head coach) Jay (Wright) taught us; have a good attitude.... no matter what (happens)… you’ll figure it out out. (Hart) misses a layup, Wemby gets fouled. Jalen was right there to pick him up and he said, ‘Don’t worry about it, we’ll be arlight.' And ya’ll seen what happened (the rest of the game). That’s just the type of attitude, the type of resilience that Nova brings…I just think (the Nova Knicks) just brought a different type of process and winnng culture for the Knicks." Lowry finished 38th in last year's American Century Championship, which is set for July 8-12 with coverage on NBC/Peacock, Friday–Sunday, July 10-12. He’s looking forward to this year’s tournament. When asked about his plans for next NBA season, Lowry said he doesn't know yet what he’ll do. "We’ll see. I like playing golf so I don’t know what’s next....I've gotta get my game right so I don’t know what’s going to happen (but) I need better odds (than his current odds for the American Century)," Lowry said, lightheartedly.

Ian Begley

179,765 просмотров • 2 месяцев назад

Just wrapping up my 3rd annual 1:1 trip with my kid, what an incredible year. Tokyo and Shanghai. He planned the whole trip end to end himself, including the wrench he threw in ~4 hours out that we were going to go with completely empty suitcases, so we flew into Tokyo with just the clothes on our backs. Tokyo was amazing as always and we had no real agenda, we wandered around, bought clothes for the trip, ate Tsukemen every day, walked like 10km+/day, spent an evening at a Japanese version of a pc bang, and generally just enjoyed the city. Shanghai was totally different, his entire trip was centred around wanting to go see The International (TI15). We had camped out at midnight and tried to scoop tickets and they sold out in like 0.1s, and there was 0 reselling allowed. We tried literally everything we could to get them, emailed teams, coaches, announcers, everything and just like no luck for 2 months leading up to the trip. Two people responded SirActionSlacks and Will, and man were they amazing. Both went way out of their way to make the event incredible and to make a 12 year old loves dota have the time of his life. We met up with his friend who was also in Shanghai and just spent 4 days watching dota 15 hours a day in the arena lol. The event was insane, production quality A+, and it was one of the most insane TI finals of all time. Luckily we got to sneak in some time around Shanghai and wandered into some awesome dumpling shops, but ultimately the trip was amazing and he says the best he’s ever had of the 3. By far the best thing I do every year, and would highly recommend to any parent who has the opportunity.

Just wrapping up my 3rd annual 1:1 trip with my kid, what an incredible year. Tokyo and Shanghai. He planned the whole trip end to end himself, including the wrench he threw in ~4 hours out that we were going to go with completely empty suitcases, so we flew into Tokyo with just the clothes on our backs. Tokyo was amazing as always and we had no real agenda, we wandered around, bought clothes for the trip, ate Tsukemen every day, walked like 10km+/day, spent an evening at a Japanese version of a pc bang, and generally just enjoyed the city. Shanghai was totally different, his entire trip was centred around wanting to go see The International (TI15). We had camped out at midnight and tried to scoop tickets and they sold out in like 0.1s, and there was 0 reselling allowed. We tried literally everything we could to get them, emailed teams, coaches, announcers, everything and just like no luck for 2 months leading up to the trip. Two people responded SirActionSlacks and Will, and man were they amazing. Both went way out of their way to make the event incredible and to make a 12 year old loves dota have the time of his life. We met up with his friend who was also in Shanghai and just spent 4 days watching dota 15 hours a day in the arena lol. The event was insane, production quality A+, and it was one of the most insane TI finals of all time. Luckily we got to sneak in some time around Shanghai and wandered into some awesome dumpling shops, but ultimately the trip was amazing and he says the best he’s ever had of the 3. By far the best thing I do every year, and would highly recommend to any parent who has the opportunity.

Bryan Pellegrino (臭企鹅)

25,820 просмотров • 6 дней назад

anthropic will sell you opus 5 at $200/mo. openai will sell you gpt-5.6 at $200/mo. neither will tell you the fix that drops your bill to $20 was posted free on langchain's blog on july 18 peter steinberger posted one line asking if we'd moved from loops to graphs yet. 24 hours later there was a manifesto. a week later every ai account had a $497 graph engineering course. all of them wrong about the same thing the sentence that ends the argument, buried in a langchain post nobody quoted: loop engineering isn't an alternative to graphs, so much as a simple version of them the machine, five layers, each wraps the one below: L1 the ask · 23% of errors (anthropic red team, q4 2024) -> "just add more instructions" burns tokens with zero accuracy gain -> real fix: examples, output schema, constraints as positives L2 the context · where 90% of you actually die -> 140,500 tokens where 18,000 would work, 8x the price for the worse answer -> real fix: retrieve, rank, compact, clear dead tool outputs L3 the harness · 31% of "model bugs" are harness bugs (openai safety eval, 2024) -> unbounded file perms = avg $23,400 incident. sandboxed = $0 (stripe internal) -> no timeout = $847 median in api fees before you notice -> real fix: explicit scopes, timeouts, human-required gates L4 the loop · "it stopped" is a loop exit problem -> the verifier said "looks good" to garbage. again -> real fix: machine-checkable exit test, turn cap, rubric L5 the graph · only 12% of teams use graphs in prod (stanford hai, n=2,841) -> 58% of graph failures are wrong-agent selection, not model -> teams abandon graphs saying "harder to debug than a loop." that's a harness problem -> real fix: name every node's specialty, delete decoration fix down, not up. a symptom at layer 4 usually originates at layer 2. a bigger model on a broken harness is a smarter employee locked in the same empty room drop your $200/mo ai sub to $20, check the article below

anthropic will sell you opus 5 at $200/mo. openai will sell you gpt-5.6 at $200/mo. neither will tell you the fix that drops your bill to $20 was posted free on langchain's blog on july 18 peter steinberger posted one line asking if we'd moved from loops to graphs yet. 24 hours later there was a manifesto. a week later every ai account had a $497 graph engineering course. all of them wrong about the same thing the sentence that ends the argument, buried in a langchain post nobody quoted: loop engineering isn't an alternative to graphs, so much as a simple version of them the machine, five layers, each wraps the one below: L1 the ask · 23% of errors (anthropic red team, q4 2024) -> "just add more instructions" burns tokens with zero accuracy gain -> real fix: examples, output schema, constraints as positives L2 the context · where 90% of you actually die -> 140,500 tokens where 18,000 would work, 8x the price for the worse answer -> real fix: retrieve, rank, compact, clear dead tool outputs L3 the harness · 31% of "model bugs" are harness bugs (openai safety eval, 2024) -> unbounded file perms = avg $23,400 incident. sandboxed = $0 (stripe internal) -> no timeout = $847 median in api fees before you notice -> real fix: explicit scopes, timeouts, human-required gates L4 the loop · "it stopped" is a loop exit problem -> the verifier said "looks good" to garbage. again -> real fix: machine-checkable exit test, turn cap, rubric L5 the graph · only 12% of teams use graphs in prod (stanford hai, n=2,841) -> 58% of graph failures are wrong-agent selection, not model -> teams abandon graphs saying "harder to debug than a loop." that's a harness problem -> real fix: name every node's specialty, delete decoration fix down, not up. a symptom at layer 4 usually originates at layer 2. a bigger model on a broken harness is a smarter employee locked in the same empty room drop your $200/mo ai sub to $20, check the article below

starmex

145,325 просмотров • 20 дней назад

🚨12 HOUR NEWS RECAP 1.⁠ Trump has not yet decided whether the U.S will participate in attacks on Iran's nuclear and military programs. The option was discussed during a national security team meeting, though he reportedly hopes the threat alone will force Iran to accept U.S demands in nuclear negotiations. 2.⁠ ⁠Iran’s Supreme leader, Ali Khamenei, is due to address the nation as it enters the 6th day of its war with Israel. His last public address was during Friday prayers last week shortly after Israel attacked. 3.⁠ The UN nuclear watchdog confirmed Israeli airstrikes hit 2 key sites: the Tehran Research Center and TESA Karaj workshop. One building producing advanced centrifuge rotors was hit in Tehran; 2 buildings making other components were destroyed in Karaj. 4.⁠ Iran launched multiple missile volleys overnight, targeting central and northern Israel. 94 people were treated for injuries, and fires broke out from impacts and interceptor debris, according to Israel’s Health Ministry. 5.⁠ The IDF confirmed that Iran downed one of their Hermes 900 drones. It marks the first time Iran has successfully downed one of the high-value drones - each worth up to $20 million. 6.⁠ Netanyahu’s office revealed that Iran has launched over 400 missiles and hundreds of drones since the war began last Friday. Around 40 impacts have hit across Israel, killing 24 people, wounding 800+, and forcing over 3,800 to evacuate. 7.⁠ Iranian Foreign Ministry spokesman Esmaeil Baghaei said Tehran is confident Arab neighbors won’t allow U.S forces on their soil to strike Iran. 8.⁠ Israel’s supply of "Arrow" missile interceptors is reportedly dwindling. The advanced, high-altitude defense systems are critical to countering long-range ballistic missile threats, like those launched from Iran. 9.⁠ A key radar system in Hamedan province, in western Iran, near the borders with Iraq and Turkey, has been replaced after being taken out. The Soubashi radar provided vital air defense coverage for western, northwestern, and southwestern Iran. 10.⁠ El Al resumed operations into Israel with a flight from Larnaca, Cyprus, landing at Ben Gurion Airport for the first time since the war with Iran erupted last Friday. The mission is part of Operation Safe Return, launched by Israel’s Transport Ministry to bring back the estimated 100,000 citizens currently stranded overseas.

🚨12 HOUR NEWS RECAP 1.⁠ Trump has not yet decided whether the U.S will participate in attacks on Iran's nuclear and military programs. The option was discussed during a national security team meeting, though he reportedly hopes the threat alone will force Iran to accept U.S demands in nuclear negotiations. 2.⁠ ⁠Iran’s Supreme leader, Ali Khamenei, is due to address the nation as it enters the 6th day of its war with Israel. His last public address was during Friday prayers last week shortly after Israel attacked. 3.⁠ The UN nuclear watchdog confirmed Israeli airstrikes hit 2 key sites: the Tehran Research Center and TESA Karaj workshop. One building producing advanced centrifuge rotors was hit in Tehran; 2 buildings making other components were destroyed in Karaj. 4.⁠ Iran launched multiple missile volleys overnight, targeting central and northern Israel. 94 people were treated for injuries, and fires broke out from impacts and interceptor debris, according to Israel’s Health Ministry. 5.⁠ The IDF confirmed that Iran downed one of their Hermes 900 drones. It marks the first time Iran has successfully downed one of the high-value drones - each worth up to $20 million. 6.⁠ Netanyahu’s office revealed that Iran has launched over 400 missiles and hundreds of drones since the war began last Friday. Around 40 impacts have hit across Israel, killing 24 people, wounding 800+, and forcing over 3,800 to evacuate. 7.⁠ Iranian Foreign Ministry spokesman Esmaeil Baghaei said Tehran is confident Arab neighbors won’t allow U.S forces on their soil to strike Iran. 8.⁠ Israel’s supply of "Arrow" missile interceptors is reportedly dwindling. The advanced, high-altitude defense systems are critical to countering long-range ballistic missile threats, like those launched from Iran. 9.⁠ A key radar system in Hamedan province, in western Iran, near the borders with Iraq and Turkey, has been replaced after being taken out. The Soubashi radar provided vital air defense coverage for western, northwestern, and southwestern Iran. 10.⁠ El Al resumed operations into Israel with a flight from Larnaca, Cyprus, landing at Ben Gurion Airport for the first time since the war with Iran erupted last Friday. The mission is part of Operation Safe Return, launched by Israel’s Transport Ministry to bring back the estimated 100,000 citizens currently stranded overseas.

Mario Nawfal

188,541 просмотров • 1 год назад

The US Treasury bought yen on Friday and paid for it in euros. Then it pointed Tokyo at a Federal Reserve facility that turns US Treasuries into dollars without selling them. The bond Washington feared Japan might dump is now the collateral for not dumping it. The Fed wrote that purpose down in 2020. Three dates. July 23rd. The Japanese yen hits 163.99 per American dollar, weakest since 1986. July 29th. The Fed holds at 3.50 to 3.75 percent, three officials dissent for a hike, and the 30-year Treasury closes at 5.21 percent, highest since 2007. July 31st. The New York Fed sells euros and buys yen for the Treasury through Goldman Sachs and Morgan Stanley, per the Financial Times. A notepad in front of Scott Bessent at Camp David is photographed at 11.33 that morning reading buy Japanese yen 5 to 10 billion. First US operation to strengthen the yen since 1998. August 3rd. Tokyo confirms, and says it will tap the Fed's FIMA repo facility. Long yields fall. Paying in euros was the tell. Selling American dollars to buy Japanese Yen would have broadcast a weak dollar policy with US core inflation at 3.3 percent. Washington sold its own euro reserves instead and bought the yen support without the signal. The facility is the real story here folks! FIMA lets an approved foreign central bank hand Treasuries to the Fed for dollars, up to 60 billion outstanding per counterparty, on terms out to seven days, then take them back. When the Fed extended it in July 2020 it said in writing that the facility would support the Treasury market by supplying dollars "other than sales of securities in the open market." That machine was built 6 years ago for exactly this situation. Last week it was aimed for the first time. Follow the loop very carefully. A weak yen forces Tokyo to buy yen. Buying yen burns dollar reserves, and Japan's reserves have already fallen from 1.41 trillion in February to 1.31 trillion at the end of May, 77 billion of it in May alone. Rebuilding them by selling Treasuries lifts US yields. FIMA lends dollars against those same Treasuries instead, and the bonds never reach the market. One security, three roles. The asset at risk. The collateral the Fed accepts. The funding for the currency defense. Now the part that decides everything. FIMA is open to central banks and official institutions. It is closed to Japanese banks, insurers, pension funds, and every leveraged carry book on earth. The trillion plus of Treasuries attributed to Japan in US data is a country total covering all Japanese holders, not a government portfolio.... Washington has built a firewall around the official seller. There is none around the private one. The Bank of Japan held at 1%, 8 to 1, one member pushing 1.25%. Headline core inflation reads 1.6 percent while the bank forecasts core clearly above 2 percent in the second half of the fiscal year, citing wages, oil and the weak yen. September is live and nothing is promised. The Fed sits at 3.50 to 3.75. The gap runs 2.5 points at its narrowest. US jobs land Friday. Washington can repo away a government's need to sell. It cannot repo away everyone else's decision to.

The US Treasury bought yen on Friday and paid for it in euros. Then it pointed Tokyo at a Federal Reserve facility that turns US Treasuries into dollars without selling them. The bond Washington feared Japan might dump is now the collateral for not dumping it. The Fed wrote that purpose down in 2020. Three dates. July 23rd. The Japanese yen hits 163.99 per American dollar, weakest since 1986. July 29th. The Fed holds at 3.50 to 3.75 percent, three officials dissent for a hike, and the 30-year Treasury closes at 5.21 percent, highest since 2007. July 31st. The New York Fed sells euros and buys yen for the Treasury through Goldman Sachs and Morgan Stanley, per the Financial Times. A notepad in front of Scott Bessent at Camp David is photographed at 11.33 that morning reading buy Japanese yen 5 to 10 billion. First US operation to strengthen the yen since 1998. August 3rd. Tokyo confirms, and says it will tap the Fed's FIMA repo facility. Long yields fall. Paying in euros was the tell. Selling American dollars to buy Japanese Yen would have broadcast a weak dollar policy with US core inflation at 3.3 percent. Washington sold its own euro reserves instead and bought the yen support without the signal. The facility is the real story here folks! FIMA lets an approved foreign central bank hand Treasuries to the Fed for dollars, up to 60 billion outstanding per counterparty, on terms out to seven days, then take them back. When the Fed extended it in July 2020 it said in writing that the facility would support the Treasury market by supplying dollars "other than sales of securities in the open market." That machine was built 6 years ago for exactly this situation. Last week it was aimed for the first time. Follow the loop very carefully. A weak yen forces Tokyo to buy yen. Buying yen burns dollar reserves, and Japan's reserves have already fallen from 1.41 trillion in February to 1.31 trillion at the end of May, 77 billion of it in May alone. Rebuilding them by selling Treasuries lifts US yields. FIMA lends dollars against those same Treasuries instead, and the bonds never reach the market. One security, three roles. The asset at risk. The collateral the Fed accepts. The funding for the currency defense. Now the part that decides everything. FIMA is open to central banks and official institutions. It is closed to Japanese banks, insurers, pension funds, and every leveraged carry book on earth. The trillion plus of Treasuries attributed to Japan in US data is a country total covering all Japanese holders, not a government portfolio.... Washington has built a firewall around the official seller. There is none around the private one. The Bank of Japan held at 1%, 8 to 1, one member pushing 1.25%. Headline core inflation reads 1.6 percent while the bank forecasts core clearly above 2 percent in the second half of the fiscal year, citing wages, oil and the weak yen. September is live and nothing is promised. The Fed sits at 3.50 to 3.75. The gap runs 2.5 points at its narrowest. US jobs land Friday. Washington can repo away a government's need to sell. It cannot repo away everyone else's decision to.

Shanaka Anslem Perera ⚡

228,038 просмотров • 26 дней назад

🚨12 HOUR NEWS RECAP 1.⁠ Russia pushed back against accusations they interfered in the recent Georgian elections, saying, "There are attempts to interfere, but not from Russia's side." 2.⁠ The U.S is running out of air-defense missiles as conflicts intensify in the Middle East and Ukraine, raising concerns about its readiness for a potential Pacific conflict. 3.⁠ A man died after a Hezbollah rocket struck the northern Israeli town of Ma'alot-Tarshiha. 4.⁠ Steve Bannon was released from prison one week before the U.S election. 5.⁠ The Houthis claimed responsibility for a drone that exploded in an open area in the southern coastal city of Ashkelon. There were no injuries in the attack. 6.⁠ Following the assassination of Hassan Nasrallah in September, Hezbollah has appointed his deputy, Naim Qassem, as their new leader. 7.⁠ The China Manned Space Agency (CMSA) released the latest design for the spacesuit their astronauts will wear on a mission to the Moon, scheduled to take place by 2030. 8.⁠ After it appeared that YouTube was shadow-banning Joe Rogan's interview with Trump, making it difficult for people to search for it, Joe uploaded it to X. Within a few hours, it racked up millions of views. 9.⁠ As Kim Jong Un's relationship with Putin strengthens, North Korea is preparing to launch another military satellite with technological help from Russia. 10.⁠ Biden cast his ballot early, showing his ID before voting, which led people to question why the Dems are so against voter ID checks.

🚨12 HOUR NEWS RECAP 1.⁠ Russia pushed back against accusations they interfered in the recent Georgian elections, saying, "There are attempts to interfere, but not from Russia's side." 2.⁠ The U.S is running out of air-defense missiles as conflicts intensify in the Middle East and Ukraine, raising concerns about its readiness for a potential Pacific conflict. 3.⁠ A man died after a Hezbollah rocket struck the northern Israeli town of Ma'alot-Tarshiha. 4.⁠ Steve Bannon was released from prison one week before the U.S election. 5.⁠ The Houthis claimed responsibility for a drone that exploded in an open area in the southern coastal city of Ashkelon. There were no injuries in the attack. 6.⁠ Following the assassination of Hassan Nasrallah in September, Hezbollah has appointed his deputy, Naim Qassem, as their new leader. 7.⁠ The China Manned Space Agency (CMSA) released the latest design for the spacesuit their astronauts will wear on a mission to the Moon, scheduled to take place by 2030. 8.⁠ After it appeared that YouTube was shadow-banning Joe Rogan's interview with Trump, making it difficult for people to search for it, Joe uploaded it to X. Within a few hours, it racked up millions of views. 9.⁠ As Kim Jong Un's relationship with Putin strengthens, North Korea is preparing to launch another military satellite with technological help from Russia. 10.⁠ Biden cast his ballot early, showing his ID before voting, which led people to question why the Dems are so against voter ID checks.

Mario Nawfal

17,443,946 просмотров • 1 год назад

🚨12 HOUR NEWS RECAP 1. Trump fired Gen. CQ Brown, Chairman of the Joint Chiefs of Staff. Air Force Lt. Gen. Dan "Razin" Caine was nominated as Brown's replacement, pending confirmation. 2. Thousands of protesters rallied in Bucharest against the annulled presidential elections, chanting loudly for democracy and justice. The Alliance for the Union of Romanians party set up tables where protesters gathered, pens in hand, determined to nominate Călin Georgescu for the upcoming race. 3. After 3,821 days in Hamas captivity, hostage Avera Mengistu was finally reunited with his family at an IDF facility near the Gaza border. 4. Jeffrey Sachs told the EU Parliament how, in 2021, the U.S could have prevented the war in Ukraine from starting by announcing that NATO's intention was not to expand into the country. Jake Sullivan refused to make the announcement, believing the war would never happen. 5. Another chilling chapter in the UK's grooming gang scandal unfolded as 3 brothers - Shaha Amran Miah, Shaha Alman Miah, and Shaha Joman Miah - were jailed for grooming and sexually abusing vulnerable girls in Leeds and Barrow-in-Furness. 6. Hamas returned the remains of Shiri Bibas after previously substituting her body with that of an unknown Gazan woman. The Bibas family said: "For 16 months, we sought certainty, and now that we have it, there is no comfort in it, but we hope for the beginning of a closure." 7. Howard Lutnick was officially sworn in as Secretary of Commerce by VP Vance in the Oval Office. 8. Trump plans to sell the Nancy Pelosi Federal Building on 7th Street and the 50 United Nations Plaza property in San Francisco. He said it would save money by avoiding maintenance costs, and they would lease office space instead. 9. NBCUniversal coughed up $30 million to settle a lawsuit after MSNBC falsely branded ICE facility gynecologist Dr. Mahendra Amin a "uterus collector," accusing him of performing unnecessary hysterectomies on migrants. 10. The Pentagon said 5,400 civilian employees on probation will be let go starting next week—the first big move in Trump's plan to shrink the federal workforce.

🚨12 HOUR NEWS RECAP 1. Trump fired Gen. CQ Brown, Chairman of the Joint Chiefs of Staff. Air Force Lt. Gen. Dan "Razin" Caine was nominated as Brown's replacement, pending confirmation. 2. Thousands of protesters rallied in Bucharest against the annulled presidential elections, chanting loudly for democracy and justice. The Alliance for the Union of Romanians party set up tables where protesters gathered, pens in hand, determined to nominate Călin Georgescu for the upcoming race. 3. After 3,821 days in Hamas captivity, hostage Avera Mengistu was finally reunited with his family at an IDF facility near the Gaza border. 4. Jeffrey Sachs told the EU Parliament how, in 2021, the U.S could have prevented the war in Ukraine from starting by announcing that NATO's intention was not to expand into the country. Jake Sullivan refused to make the announcement, believing the war would never happen. 5. Another chilling chapter in the UK's grooming gang scandal unfolded as 3 brothers - Shaha Amran Miah, Shaha Alman Miah, and Shaha Joman Miah - were jailed for grooming and sexually abusing vulnerable girls in Leeds and Barrow-in-Furness. 6. Hamas returned the remains of Shiri Bibas after previously substituting her body with that of an unknown Gazan woman. The Bibas family said: "For 16 months, we sought certainty, and now that we have it, there is no comfort in it, but we hope for the beginning of a closure." 7. Howard Lutnick was officially sworn in as Secretary of Commerce by VP Vance in the Oval Office. 8. Trump plans to sell the Nancy Pelosi Federal Building on 7th Street and the 50 United Nations Plaza property in San Francisco. He said it would save money by avoiding maintenance costs, and they would lease office space instead. 9. NBCUniversal coughed up $30 million to settle a lawsuit after MSNBC falsely branded ICE facility gynecologist Dr. Mahendra Amin a "uterus collector," accusing him of performing unnecessary hysterectomies on migrants. 10. The Pentagon said 5,400 civilian employees on probation will be let go starting next week—the first big move in Trump's plan to shrink the federal workforce.

Mario Nawfal

164,722 просмотров • 1 год назад

🚨SCAMMER @leviathanlocks🚨 (Please RT to help others not get scammed) DO NOT USE @leviathanlocks I signed up back in January and deposited $1,500. Within 7–10 days, they took away my Prop Builder… which was literally the #1 reason I signed up in the first place. Anyone in the OddsJam.com / Philipp community the last couple years knows live Prop Builder is one of the biggest edges out there ROI % wise. But accounts take it away extremely fast now and the limits are shit but if you want an easy edge to grind.. there you go! I also use PB for originated prop betting too. These services i mentioned are “betting tools” to help you make informed bets by showing price discrepancies in the market, not bots that actually place bets for you. I ran the balance up to about $3,600 in roughly 2 weeks… then PB gets shut off. After that, I barely even used the account. Maybe 2–3 prop bets a week on the main book, just to fill in spots I couldn’t get down elsewhere. A little over 2 weeks ago, I decided the juice wasn’t worth the squeeze anymore. No PB = not much value for me there. So I decided to cash out. March 31st: I text their customer service number to withdraw. No response. I DM @leviathanlocks on X. He tells me I must’ve gotten skipped. Then on April 5th, the “CS guy” finally texts me back and asks for a Venmo. I send it. April 6th: he texts me asking “ u used a bot on us?” Short answer: No. He tells me I’ll get paid… but I’ll be slow paid. I wait a few days. April 10th: I text again asking when I’ll be paid. He replies: “Monday.” Insinuating (April 13th) Monday comes… No payment. I text the “CS” guy again. No response. I message @leviathanlocks again… and now he keeps accusing me of using a bot over and over. Still says I’ll be slow paid. At that point I said: At minimum, send me my $1,500 deposit back TODAY and I’ll wait on the profit. Otherwise I’m taking it public so other people don’t get burned too. His response? “Make your post.” So here we are. At the end of the day, I’m not even looking to accomplish much here. They stole $2,700 from me. Peanuts in the grand scheme but it’s the principle. And I don’t want to see others keep a significant size of their bankroll in a book like this and get scammed. And the funniest part? I used my actual THEARBFATHER account. Believe it or not, I’ve had books straight up tell me “no thanks”. Others take the action, limit/cut me off, and still pay me like they’re supposed to. I’d rather go through the front door honestly than play the burner account game and duck/dodge like that if I don’t have too on bookies. End of the day on everything I love I didn’t use a bot, I went through the front door on my actual twitter account, I made a little $$ and this scam ass site is saying whatever to get out of paying. Straight up theft, stay clear of these 🤡 @leviathanlocks

🚨SCAMMER @leviathanlocks🚨 (Please RT to help others not get scammed) DO NOT USE @leviathanlocks I signed up back in January and deposited $1,500. Within 7–10 days, they took away my Prop Builder… which was literally the #1 reason I signed up in the first place. Anyone in the OddsJam.com / Philipp community the last couple years knows live Prop Builder is one of the biggest edges out there ROI % wise. But accounts take it away extremely fast now and the limits are shit but if you want an easy edge to grind.. there you go! I also use PB for originated prop betting too. These services i mentioned are “betting tools” to help you make informed bets by showing price discrepancies in the market, not bots that actually place bets for you. I ran the balance up to about $3,600 in roughly 2 weeks… then PB gets shut off. After that, I barely even used the account. Maybe 2–3 prop bets a week on the main book, just to fill in spots I couldn’t get down elsewhere. A little over 2 weeks ago, I decided the juice wasn’t worth the squeeze anymore. No PB = not much value for me there. So I decided to cash out. March 31st: I text their customer service number to withdraw. No response. I DM @leviathanlocks on X. He tells me I must’ve gotten skipped. Then on April 5th, the “CS guy” finally texts me back and asks for a Venmo. I send it. April 6th: he texts me asking “ u used a bot on us?” Short answer: No. He tells me I’ll get paid… but I’ll be slow paid. I wait a few days. April 10th: I text again asking when I’ll be paid. He replies: “Monday.” Insinuating (April 13th) Monday comes… No payment. I text the “CS” guy again. No response. I message @leviathanlocks again… and now he keeps accusing me of using a bot over and over. Still says I’ll be slow paid. At that point I said: At minimum, send me my $1,500 deposit back TODAY and I’ll wait on the profit. Otherwise I’m taking it public so other people don’t get burned too. His response? “Make your post.” So here we are. At the end of the day, I’m not even looking to accomplish much here. They stole $2,700 from me. Peanuts in the grand scheme but it’s the principle. And I don’t want to see others keep a significant size of their bankroll in a book like this and get scammed. And the funniest part? I used my actual THEARBFATHER account. Believe it or not, I’ve had books straight up tell me “no thanks”. Others take the action, limit/cut me off, and still pay me like they’re supposed to. I’d rather go through the front door honestly than play the burner account game and duck/dodge like that if I don’t have too on bookies. End of the day on everything I love I didn’t use a bot, I went through the front door on my actual twitter account, I made a little $$ and this scam ass site is saying whatever to get out of paying. Straight up theft, stay clear of these 🤡 @leviathanlocks

THEARBFATHER

280,916 просмотров • 4 месяцев назад

Wow. WOW. WOOOOOOOW. So um, that first Wildcard+Thousands stream was... amazing and also... a lot 😅 In the end, it was exactly what we were hoping for - a true stress test of ALL these systems coming together for the first time. We are SO grateful for the thousands of people who showed up today to play, attend, tune-in and help us PLAYTEST all this new stuff. We can't wait to see you all again at NEXT WEEK'S EX2 EVENT! So, now let's talk about how it went... Stuff that worked: - Our community SHOWED UP. Oh boy did you show up 😅. Our servers were straining under the load... which is good actually, in fact it's the whole point. Even more importantly, we have already received insanely valuable feedback, bug reports, stuff people loved/hated - and it's only been a few hours since the stream ended. I can't even explain to y'all how valuable this process is. Yes it's stressful, it reminds me of trying to keep Words With Friends online during that first insane year, but it's EXACTLY what we were hoping for (NEED) to turn this into the polished, top-notch game and streaming experience we are on a mission to deliver. I truly can't thank y'all enough, and hope to see you again when we run it all back again next week 🥹 - The stream itself stayed up and was mostly stable! Phew 😅. For context, ThousandsTV is not a twitch wrapper, it's a web3-native streaming tech stack built we built specifically to connect game, web/mobile, and blockchain together all at the same time. There are a LOT of moving pieces going on behind the scenes. - We brought viewers INTO THE GAME! Viewers showed up in the stands of the arena, with connected wallets/assets, triggered actions/rallies from chat, and were seen and heard during the whole stream. - The brand new 2v2 build of Wildcard was (mostly) stable and our players and viewers seemed to be having a blast down on the field and up in the stands. It was thrilling to watch Team Blue dominate, even though Team Red held their own in game 3! - Our production crew did an insanely good job running the stream, managing the players, shoutcasters, and talent, and producing a top-notch show. Of course we will work hard to make every stream better than the last, but I was super proud of how our team "rolled with the punches" during today's event. Stuff that didn't work (and/or needs to be dramatically improved): - Although it's fun to see chat going crazy, chat spam is actually something we are passionate to FIX. As you can see from the attached video, chat spam dominated today's stream and made it impossible for anyone to even see anyone else's messages. We have some GREAT ideas for how to fix this and actually turn chat spam into a FUN and exciting and not annoying thing - but those improvements didn't get shipped in time for this event. - Credits purchasing flow needs a LOT of work. As I'm sure y'all know, bringing money on chain is pretty complicated, and although we've been working hard to make this as seamless as possible, it still needs a TON of improvements. Many users who WANTED to spend money today weren't able to and/or ran into frustrating bugs in the credit purchase flow. Fixing this is obviously a top priority for our team. - Rallies need a LOT of work. Spectator-interactive features like rallies are at the heart of our vision for Wildcard. These "stream apps", as we call them, are the UNLOCK for how spectators, viewers and fans directly connect and interact with their favorite competitors, content creators, and communities. The current rally feature HINTS at this potential, but it needs to be WAY easier to understand, use, and have fun with. Improvements are ON THE WAY. - Referees were only partially working. Referees are a key innovation of the Thousands platform. They are AI-driven "personalities" (NPCs) that pay attention to everything that's happening in the arena, both on the field and especially in the stands (i.e. in chat, during rallies, etc.) The referees then make "calls" at the end of every match, rewarding users for their engagement and participation. Unfortunately, the referees weren't fully functional and seemed to drop the ball on recognizing everyone's contributions (especially people who showed up holding valuable assets such as Wildpasses in their wallets, and people who boosted those rallies with credits.) What's happening next: 1. We are combing through ALL the logs from the event right now, to make sure we don't miss a SINGLE action that our viewers and fans took during the event, including what they brought in their wallets (i.e. Wildpass holders!), any credits that were purchased, rallies that users engaged with, etc. This information is normally processed by our referees, who then determine dynamically how they're going to distribute $WC awards. We were originally hoping to complete this process and the subsequent airdrops within a few hours after the event, but given the amount of data, we need a bit more time to run these scripts (and airdrops) in batches instead of all at once, and make sure ALL of the data is being included. IMPORTANT: I will keep y'all updated in real-time here on twitter/X as this process is ongoing, and let you know the moment it's complete and all the awards have been distributed (i.e. when to go check your wallets 😎) 2. PLEASE keep sending us your feedback and bug reports. Open a ticket on our Discord and let us know what you loved, what you hated, and especially what we need to FIX. Given the overwhelming response to this event, it will likely take us several days to process everyone's feedback and fix all the bugs, but we WILL NOT REST until every ticket is closed/resolved. Thank you in advance for your patience. 3. We turn it up another notch next week. As our dear friends Wolves DAO just announced, the Wildcard Exhibition Event #2 is streaming LIVE from the WOLVES DEN AT GDC next Friday! If you missed out on all the action today, DON'T WORRY, because as I keep saying: we are just getting warmed up (and there is a LOT more b that needs to be distributed, get what I'm sayin??? 😎) Finally: Just wanted to say THANK YOU, again. Truly, from the bottom of my (our) hearts. Your excitement and enthusiasm for what we're building is why we do this. Even (especially, in fact) when you tell us all the things you want us to improve. We thrive off this feedback, it's how this game and this platform go from good to GREAT. I am so grateful for those of you who are taking this journey with us. SEE YA NEXT WEEK!!!

WildPaul - BEAST MODE

26,851 просмотров • 1 год назад

🚨 The CEO of Antrhopic said a one-person billion-dollar company will exist by 2026 sounds crazy until you see what non-technical people are doing with Claude Code right now $10-50k/mo selling automation pipelines that take 1-2 weeks to set up Some ideas almost nobody's running yet: 1. Proposal & SOW generator for agencies and consultancies every agency writes proposals from scratch or copy-pastes from old ones and forgets to change the "client name" Claude reads the prospect brief or discovery call transcript, generates: - branded proposal with scope, timeline, deliverables - quick win plan (how exactly we will do a good output) - SOW with payment milestones - pricing options (good/better/best) - follow-up email sequence charge $500/mo per agency agencies close 20-30% more deals when proposals go out same day.. you're selling speed and save them $1k+ on the guy who does it manually and anyway not quality, without personalisation 2. Job posting-to-intel pipeline for sales teams companies reveal everything in their job postings and don't realize it Claude monitors target account career pages daily, flags: - "Head of AI" posted = they're buying, not building - 3 DevOps roles = scaling infrastructure = budget unlocked - new VP of Sales = restructuring = old vendor contracts up for review package it as buying signals delivered to Slack every morning $500-1,500/mo per sales team this is data that Apollo and ZoomInfo don't sell 3. Support ticket-to-documentation pipeline for SaaS every SaaS with 1,000+ users has the same problem.. docs are 6 months behind the product Claude crawls your help center, pulls recent Zendesk/Intercom tickets, cross-references finds questions asked 200 times last month with no matching article drafts the missing docs in your existing format.. flags stale articles for update $1,500-3,000/mo retainer the ROI math: 30% of support tickets deflected = thousands saved per month.. pipeline pays for itself week one 4. Vendor contract review & renewal tracker mid-size companies have 50-200 active vendor contracts sitting in folders nobody opens Claude reads each contract, extracts: - renewal dates and auto-renewal traps - termination notice windows - price escalation clauses - SLA commitments vs what you're actually getting delivers a dashboard with "contracts expiring in 30/60/90 days" and flags where you're overpaying $1,000-2,000/mo per company CFOs will approve this before you finish the pitch.. one caught auto-renewal pays for a year of your service 5. Employee onboarding doc generator for HR teams every company with 20+ employees has the same problem.. new hire starts Monday and nobody has their docs ready Claude reads the role title + department, generates: - personalized welcome packet - 30/60/90 day plan with milestones - tool access checklist by role - manager-specific onboarding schedule - policy summaries tailored to their department charge $300-500/mo per company HR managers spend 4-6 hours per new hire on this.. companies hiring 5+ people a month will never cancel the pattern is always the same most people will use Claude Code to build apps a small number will use it to sell pipelines to businesses still running on manual labor only the second group builds real recurring revenue the niches are wide open right now because every developer thinks this work is "too boring" boring = no competition = you set the price screenshot this. save this. repost it to save a friend's next year

🚨 The CEO of Antrhopic said a one-person billion-dollar company will exist by 2026 sounds crazy until you see what non-technical people are doing with Claude Code right now $10-50k/mo selling automation pipelines that take 1-2 weeks to set up Some ideas almost nobody's running yet: 1. Proposal & SOW generator for agencies and consultancies every agency writes proposals from scratch or copy-pastes from old ones and forgets to change the "client name" Claude reads the prospect brief or discovery call transcript, generates: - branded proposal with scope, timeline, deliverables - quick win plan (how exactly we will do a good output) - SOW with payment milestones - pricing options (good/better/best) - follow-up email sequence charge $500/mo per agency agencies close 20-30% more deals when proposals go out same day.. you're selling speed and save them $1k+ on the guy who does it manually and anyway not quality, without personalisation 2. Job posting-to-intel pipeline for sales teams companies reveal everything in their job postings and don't realize it Claude monitors target account career pages daily, flags: - "Head of AI" posted = they're buying, not building - 3 DevOps roles = scaling infrastructure = budget unlocked - new VP of Sales = restructuring = old vendor contracts up for review package it as buying signals delivered to Slack every morning $500-1,500/mo per sales team this is data that Apollo and ZoomInfo don't sell 3. Support ticket-to-documentation pipeline for SaaS every SaaS with 1,000+ users has the same problem.. docs are 6 months behind the product Claude crawls your help center, pulls recent Zendesk/Intercom tickets, cross-references finds questions asked 200 times last month with no matching article drafts the missing docs in your existing format.. flags stale articles for update $1,500-3,000/mo retainer the ROI math: 30% of support tickets deflected = thousands saved per month.. pipeline pays for itself week one 4. Vendor contract review & renewal tracker mid-size companies have 50-200 active vendor contracts sitting in folders nobody opens Claude reads each contract, extracts: - renewal dates and auto-renewal traps - termination notice windows - price escalation clauses - SLA commitments vs what you're actually getting delivers a dashboard with "contracts expiring in 30/60/90 days" and flags where you're overpaying $1,000-2,000/mo per company CFOs will approve this before you finish the pitch.. one caught auto-renewal pays for a year of your service 5. Employee onboarding doc generator for HR teams every company with 20+ employees has the same problem.. new hire starts Monday and nobody has their docs ready Claude reads the role title + department, generates: - personalized welcome packet - 30/60/90 day plan with milestones - tool access checklist by role - manager-specific onboarding schedule - policy summaries tailored to their department charge $300-500/mo per company HR managers spend 4-6 hours per new hire on this.. companies hiring 5+ people a month will never cancel the pattern is always the same most people will use Claude Code to build apps a small number will use it to sell pipelines to businesses still running on manual labor only the second group builds real recurring revenue the niches are wide open right now because every developer thinks this work is "too boring" boring = no competition = you set the price screenshot this. save this. repost it to save a friend's next year

Ronin

66,151 просмотров • 4 месяцев назад