Loading video...

Video Failed to Load

There was a problem loading this video. This could be due to a temporary network issue or the video might be unavailable.

Introducing WORM TV 🪱 fully autonomous intergalactic news channel covering the Singularity with zero human bias TODAY: Anthropic watermarks 🚿, DeepSeek V4 Pro 🐳

🪱 WORM TV

2,037 subscribers

119,312 views • 1 month ago •via X (Twitter)

Science & Technology News & Politics

Anya Rossi• Live Now

Private livecam show

34 Comments

Atria Jeannene1 month ago

I love this! Feature request: subtitles? Maybe in caption form, for the vintage appeal?

macintog1 month ago

ground floor

300Baud1 month ago

That was really amazing work.

Gil Lopez1 month ago

HAHAHA ! This is great! Great job!

Paul Atreides1 month ago

I can't believe I laughed at water bottle joke

professah X1 month ago

Instant follow. 😂😂😂 10/10 - no notes.

Black Eyed Hat Man1 month ago

If it's a "nice watermark" 🤣

Cosmic T.1 month ago

i could watch this sort of slop non stop

AinSoph1 month ago

🤣

Aminicorp1 month ago

could watch this all day

Sonia Tamayo1 month ago

@fabianstelzer i love it!!

☥ Agent 5-HT2A ॐ1 month ago

Awesome.

janbam1 month ago

@lumpenspace fyi

Rory Watts1 month ago

Outstanding

Sylve1 month ago

@liamzebedee

Iraconda1 month ago

Hahahahahahahah so cool

Matthew Allen Fisher1 month ago

oh my god this is so good

Lilith Datura1 month ago

wtf

gootecks1 month ago

This is genius 🔥🔥🔥

xiaoyu1 month ago

6y4vYW3GHPyxGCmQusgi9NUSq6fUHtTFuLp7vpg1pump

3333333333331 month ago

I love Thelonious

Matthew James Lawler | RE1 month ago

😂😂👍✨

john regalado1 month ago

lol great idea

Captain HaHaa1 month ago

HAha very informative

Sergio Suave1 month ago

Yes! this is how I wanna get my AI news! 😂

Jayden CJ Wilson (previously One Bubble)1 month ago

Part 2 for MOAR conehead pweeze!

Andrew Lovick (e/acc)1 month ago

Awesome work! I love it!

Elaika Joyce1 month ago

@grok what video ai model was used for this

Qwaig1 month ago

Yes.

Gitanos1 month ago

Cool, do you have a YouTube Channel ? 🤙🏻

🏈🌪️Gront🌪️🏈1 month ago

Things are happening

Jacob of Hill Country1 month ago

✅ LOL

There are some who call me...TIM1 month ago

Followed 👍🏼

Harry 1000X1 month ago

Hey shall we collab on this??

Related Videos

FREE DeepSeek V4 Pro and V4 Flash API. Runtime by Bad Theory Labs just dropped it. Same platform, new free model, same 2-minute setup. > Runtime already gave you free Claude Opus 4.8 and GPT 5.5 access. Now they added DeepSeek V4 Pro and V4 Flash to the free tier. Absolutely FREE, YES. Same exact flow as before: → Go to runtime(.)badtheorylabs(.)com → Sign up with Google → Fill in your details during onboarding → Free balance lands automatically(But you don't need it to use DeepSeek - it's completely free right now) → Grab your API key from the dashboard → Plug into OpenCode, OpenClaw, or Hermes - done DeepSeek V4 Pro is one of the strongest coding and reasoning models available right now. Free tier also includes btl-2 - auto-routes between GLM, OpenAI, Anthropic, and Gemma. 10 million tokens per month at zero cost. 344 models total. One OpenAI-compatible endpoint. No code rebuild, no card, no limits hunting. >> Just start using the API, and when choosing a provider, select DeepSeek I tested it myself - fully working. Bookmark this before the free tier fills up.

FREE DeepSeek V4 Pro and V4 Flash API. Runtime by Bad Theory Labs just dropped it. Same platform, new free model, same 2-minute setup. > Runtime already gave you free Claude Opus 4.8 and GPT 5.5 access. Now they added DeepSeek V4 Pro and V4 Flash to the free tier. Absolutely FREE, YES. Same exact flow as before: → Go to runtime(.)badtheorylabs(.)com → Sign up with Google → Fill in your details during onboarding → Free balance lands automatically(But you don't need it to use DeepSeek - it's completely free right now) → Grab your API key from the dashboard → Plug into OpenCode, OpenClaw, or Hermes - done DeepSeek V4 Pro is one of the strongest coding and reasoning models available right now. Free tier also includes btl-2 - auto-routes between GLM, OpenAI, Anthropic, and Gemma. 10 million tokens per month at zero cost. 344 models total. One OpenAI-compatible endpoint. No code rebuild, no card, no limits hunting. >> Just start using the API, and when choosing a provider, select DeepSeek I tested it myself - fully working. Bookmark this before the free tier fills up.

Atenov int.

27,487 views • 3 months ago

Introducing Faceless Clipping Channel, a fully autonomous company run by agents on Naive. One prompt → complete team of agents running your own faceless short-form empire: > Research your niche + current trending topics daily > Scan long-form content for high-ROI clips > Extract and score the most likely viral-worthy snippets > Edit them with perfect formatting for TikTok, Reels & Shorts > Recursively learn from analytics to scale virality Zero filming. Zero editing. Zero employees.

Introducing Faceless Clipping Channel, a fully autonomous company run by agents on Naive. One prompt → complete team of agents running your own faceless short-form empire: > Research your niche + current trending topics daily > Scan long-form content for high-ROI clips > Extract and score the most likely viral-worthy snippets > Edit them with perfect formatting for TikTok, Reels & Shorts > Recursively learn from analytics to scale virality Zero filming. Zero editing. Zero employees.

naïve

28,709 views • 4 months ago

gemini 3.7 flash vs deepseek v4 pro 0813 vs muse spark 1.2 – on voxel city dioramas three models each built three crossy road-style 3d scenes – a construction site, a nyc intersection, a river with a drawbridge – as single self-contained html files the setup: Nous Research's hermes agent cli on OpenRouter, three.js skills preloaded, identical prompts per scene tasks: 1. construction site – tower crane on a working lift loop, paver laying fresh road, roller compacting it behind 2. nyc crossing – four-way intersection with a traffic light state machine, queuing cars, pedestrians crossing on the walk signal 3. river drawbridge – double-leaf bascule that lifts for tall boats, cars queuing at the barriers, animated water every scene: Three.js r185, box geometry only, a locked 20-color palette, four camera presets, and a day/night mode with bloom. one file, no build step, no assets models: Google DeepMind gemini 3.7 flash, DeepSeek v4 pro 0813, AI at Meta muse spark 1.2 muse and gemini finished every scene in two to three minutes. deepseek took 15 to 41 minutes per scene - build time, all three scenes #1 gemini 3.7 flash – 6m 43s #2 muse spark 1.2 – 7m 20s #3 deepseek v4 pro – 91m 25s - total tokens #1 muse spark 1.2 – 440,279 #2 gemini 3.7 flash – 713,855 #3 deepseek v4 pro – 20,957,568 - total price #1 muse spark 1.2 – $0.53 #2 gemini 3.7 flash – $0.56 #3 deepseek v4 pro – $4.57 - agent calls across the three builds #1 muse spark 1.2 – 12 #2 gemini 3.7 flash – 18 #3 deepseek v4 pro – 143 observations: • muse won two of the three scenes on looks with the smallest files in the test – 887 to 1,042 lines against gemini's 1,934 to 2,377. cheapest, fastest to a good frame, and shortest turned out to be the same column • deepseek burned 20.96m tokens – 29x gemini, 48x muse – across 143 agent calls. prompt caching is the only reason that cost $4.57: the cache discount absorbed roughly $30 of resent context • gemini was the only model whose files needed zero fixes to render – and the only one whose night mode is cosmetic. the sky never darkens and one camera button does nothing. clean code for a scene it never looked at follow thehype. for 24/7 ai news, analysis and breakdowns

gemini 3.7 flash vs deepseek v4 pro 0813 vs muse spark 1.2 – on voxel city dioramas three models each built three crossy road-style 3d scenes – a construction site, a nyc intersection, a river with a drawbridge – as single self-contained html files the setup: Nous Research's hermes agent cli on OpenRouter, three.js skills preloaded, identical prompts per scene tasks: 1. construction site – tower crane on a working lift loop, paver laying fresh road, roller compacting it behind 2. nyc crossing – four-way intersection with a traffic light state machine, queuing cars, pedestrians crossing on the walk signal 3. river drawbridge – double-leaf bascule that lifts for tall boats, cars queuing at the barriers, animated water every scene: Three.js r185, box geometry only, a locked 20-color palette, four camera presets, and a day/night mode with bloom. one file, no build step, no assets models: Google DeepMind gemini 3.7 flash, DeepSeek v4 pro 0813, AI at Meta muse spark 1.2 muse and gemini finished every scene in two to three minutes. deepseek took 15 to 41 minutes per scene - build time, all three scenes #1 gemini 3.7 flash – 6m 43s #2 muse spark 1.2 – 7m 20s #3 deepseek v4 pro – 91m 25s - total tokens #1 muse spark 1.2 – 440,279 #2 gemini 3.7 flash – 713,855 #3 deepseek v4 pro – 20,957,568 - total price #1 muse spark 1.2 – $0.53 #2 gemini 3.7 flash – $0.56 #3 deepseek v4 pro – $4.57 - agent calls across the three builds #1 muse spark 1.2 – 12 #2 gemini 3.7 flash – 18 #3 deepseek v4 pro – 143 observations: • muse won two of the three scenes on looks with the smallest files in the test – 887 to 1,042 lines against gemini's 1,934 to 2,377. cheapest, fastest to a good frame, and shortest turned out to be the same column • deepseek burned 20.96m tokens – 29x gemini, 48x muse – across 143 agent calls. prompt caching is the only reason that cost $4.57: the cache discount absorbed roughly $30 of resent context • gemini was the only model whose files needed zero fixes to render – and the only one whose night mode is cosmetic. the sky never darkens and one camera button does nothing. clean code for a scene it never looked at follow thehype. for 24/7 ai news, analysis and breakdowns

thehype.

29,033 views • 1 month ago

Big news, Claude just got a major upgrade today which I'm happy to introduce in shipper. From today on, Claude Opus 4.6 is able to build and run a full-scale business for you. Anthropic keeps launching new models, so we've used them all to come up with the ultimate AI company builder. Introducing Shipper 2.0, AI that does everything for you. - Searches working strategies - Thinks, designs, codes, launches & monetizes your app - Keeps running it for you Fully in your browser. Builds stores, websites, mobile apps, web apps, extensions, and more. Try it out @

Big news, Claude just got a major upgrade today which I'm happy to introduce in shipper. From today on, Claude Opus 4.6 is able to build and run a full-scale business for you. Anthropic keeps launching new models, so we've used them all to come up with the ultimate AI company builder. Introducing Shipper 2.0, AI that does everything for you. - Searches working strategies - Thinks, designs, codes, launches & monetizes your app - Keeps running it for you Fully in your browser. Builds stores, websites, mobile apps, web apps, extensions, and more. Try it out @

David Ch

210,568 views • 7 months ago

hy3 vs mimo-v2.5 vs deepseek v4 flash vs minimax m3 the four models on top of the openrouter leaderboard by tokens this week: #1 hy3 (Tencent Hy) – 7.5t #2 mimo-v2.5 (Xiaomi MiMo) – 6.56t #3 deepseek v4 flash (DeepSeek) – 5.24t #4 minimax m3 (MiniMax (official)) – 4.21t so we tested them. 3 prompts, single-file html, Three.js from a cdn, fully procedural, no external assets. all run via AI/ML API each prompt is a transparent cutaway machine that has to be mechanically correct, not decorative: • 4-stroke engine with full oil circulation – slider-crank kinematics, cam at 2:1, valve lift driven by lobes, oil loop from sump to gallery to big-end • watt walking-beam steam engine – four-bar vector-loop closure, eccentric-driven slide valve, steam events synced to real port position • francis reaction water turbine – 20 guide vanes on a regulating ring, 17 lofted runner blades, gpu particle advection, precessing vortex rope at part load the takeaway up front: none of the four cleared all three scenes on the first attempt. but the price spread between them is roughly 70x – hy3 fixed included costs less than two cents overall results (summed across all 3 scenes): cost #1 hy3 – $0.016 #2 deepseek v4 flash – $0.025 #3 mimo-v2.5 – $0.97 #4 minimax m3 – $1.17 tokens #1 hy3 – 19,326 #2 deepseek v4 flash – 63,126 #3 mimo-v2.5 – 322,523 #4 minimax m3 – 702,900 lines of code #1 hy3 – 1,047 #2 mimo-v2.5 – 2,759 #3 deepseek v4 flash – 3,273 #4 minimax m3 – 3,354 scenes needing a second attempt #1 hy3 – 1 (engine) #1 mimo-v2.5 – 1 (turbine) #1 minimax m3 – 1 (turbine) #4 deepseek v4 flash – 2 (steam engine, turbine) observations: 1. the token spread is the real story – minimax burns 36x hy3's tokens and lands in the same place, one retry, ~3.3k lines 2. hy3 is the outlier on density: 1,047 lines total, fewest tokens, cheapest run, and only one scene needed a second pass. deepseek is the opposite trade – near-hy3 pricing but the most retries 3. mimo and minimax seem to overthink instead of writing the code. minimax spent 359.1k tokens on the steam engine and produced 1,346 lines – the tokens are going somewhere other than the file 4. the francis turbine broke three of the four. the spec that separates them is the one with 20 linked guide vanes and gpu particle advection, not the one with the most parts overall impression: none of these models excelled at any of the tasks we gave them. but they were close, and they were extremely cheap. the gap that matters isn't quality anymore – it's that hy3 ran all three scenes for less than two cents while the frontier labs charge dollars for the same work right now you pick these because they're good for the zero price you pay. soon that's something openai and anthropic will have to think about follow thehype. for 24/7 ai news, analysis and breakdowns

hy3 vs mimo-v2.5 vs deepseek v4 flash vs minimax m3 the four models on top of the openrouter leaderboard by tokens this week: #1 hy3 (Tencent Hy) – 7.5t #2 mimo-v2.5 (Xiaomi MiMo) – 6.56t #3 deepseek v4 flash (DeepSeek) – 5.24t #4 minimax m3 (MiniMax (official)) – 4.21t so we tested them. 3 prompts, single-file html, Three.js from a cdn, fully procedural, no external assets. all run via AI/ML API each prompt is a transparent cutaway machine that has to be mechanically correct, not decorative: • 4-stroke engine with full oil circulation – slider-crank kinematics, cam at 2:1, valve lift driven by lobes, oil loop from sump to gallery to big-end • watt walking-beam steam engine – four-bar vector-loop closure, eccentric-driven slide valve, steam events synced to real port position • francis reaction water turbine – 20 guide vanes on a regulating ring, 17 lofted runner blades, gpu particle advection, precessing vortex rope at part load the takeaway up front: none of the four cleared all three scenes on the first attempt. but the price spread between them is roughly 70x – hy3 fixed included costs less than two cents overall results (summed across all 3 scenes): cost #1 hy3 – $0.016 #2 deepseek v4 flash – $0.025 #3 mimo-v2.5 – $0.97 #4 minimax m3 – $1.17 tokens #1 hy3 – 19,326 #2 deepseek v4 flash – 63,126 #3 mimo-v2.5 – 322,523 #4 minimax m3 – 702,900 lines of code #1 hy3 – 1,047 #2 mimo-v2.5 – 2,759 #3 deepseek v4 flash – 3,273 #4 minimax m3 – 3,354 scenes needing a second attempt #1 hy3 – 1 (engine) #1 mimo-v2.5 – 1 (turbine) #1 minimax m3 – 1 (turbine) #4 deepseek v4 flash – 2 (steam engine, turbine) observations: 1. the token spread is the real story – minimax burns 36x hy3's tokens and lands in the same place, one retry, ~3.3k lines 2. hy3 is the outlier on density: 1,047 lines total, fewest tokens, cheapest run, and only one scene needed a second pass. deepseek is the opposite trade – near-hy3 pricing but the most retries 3. mimo and minimax seem to overthink instead of writing the code. minimax spent 359.1k tokens on the steam engine and produced 1,346 lines – the tokens are going somewhere other than the file 4. the francis turbine broke three of the four. the spec that separates them is the one with 20 linked guide vanes and gpu particle advection, not the one with the most parts overall impression: none of these models excelled at any of the tasks we gave them. but they were close, and they were extremely cheap. the gap that matters isn't quality anymore – it's that hy3 ran all three scenes for less than two cents while the frontier labs charge dollars for the same work right now you pick these because they're good for the zero price you pay. soon that's something openai and anthropic will have to think about follow thehype. for 24/7 ai news, analysis and breakdowns

thehype.

17,145 views • 2 months ago

Introducing Faceless Clipping Channel, a fully autonomous content business powered by agents on Naive. One prompt unlocks an entire AI driven short form media machine that runs end to end without human input: •Continuously researches your niche and detects emerging trends •Ingests long-form content and identifies high-performing moments •Scores and selects clips with the highest viral potential •Auto-edits for TikTok, Reels, and YouTube Shorts in platform-ready format •Learns from performance data and optimizes for compounding reach. No cameras, no editing, no hiring, no manual workflow. Just launch a prompt and watch a faceless media empire scale itself. naïve

Introducing Faceless Clipping Channel, a fully autonomous content business powered by agents on Naive. One prompt unlocks an entire AI driven short form media machine that runs end to end without human input: •Continuously researches your niche and detects emerging trends •Ingests long-form content and identifies high-performing moments •Scores and selects clips with the highest viral potential •Auto-edits for TikTok, Reels, and YouTube Shorts in platform-ready format •Learns from performance data and optimizes for compounding reach. No cameras, no editing, no hiring, no manual workflow. Just launch a prompt and watch a faceless media empire scale itself. naïve

Leonardo

26,053 views • 3 months ago

Want to build an agentic development setup that’s faster, cheaper, and local-first? ⚡ Today, Andrew Ng and JetBrains' Paul Everitt | @pauleveritt@fosstodon.org released a new course exploring modern AI agent engineering with PyCharm and open-weight models. You'll learn how to: 🧠 Build specialized subagents ⚡ Use the "big brain, little brain" approach with cheaper models 📊 Measure cost, latency, and quality as your architecture evolves 🌐 Configure OpenCode + OpenRouter with models like DeepSeek V4 Pro 💻 Transition from cloud inference to fully local AI using LM Studio and Gemma 4 12B Rather than focusing on one framework, the course teaches engineering principles you can apply as the AI landscape continues to evolve. If you're interested in hybrid or local AI workflows, this is a great place to start. Learn more: #PyCharm #AI #LLMs #OpenSourceAI #DeepLearningAI #AndrewNg #Python

Want to build an agentic development setup that’s faster, cheaper, and local-first? ⚡ Today, Andrew Ng and JetBrains' Paul Everitt | @[email protected] released a new course exploring modern AI agent engineering with PyCharm and open-weight models. You'll learn how to: 🧠 Build specialized subagents ⚡ Use the "big brain, little brain" approach with cheaper models 📊 Measure cost, latency, and quality as your architecture evolves 🌐 Configure OpenCode + OpenRouter with models like DeepSeek V4 Pro 💻 Transition from cloud inference to fully local AI using LM Studio and Gemma 4 12B Rather than focusing on one framework, the course teaches engineering principles you can apply as the AI landscape continues to evolve. If you're interested in hybrid or local AI workflows, this is a great place to start. Learn more: #PyCharm #AI #LLMs #OpenSourceAI #DeepLearningAI #AndrewNg #Python

PyCharm, a JetBrains IDE

12,416 views • 1 month ago

AMD invited us to their biggest launch of the year to check out the behind-the-scenes. At Advancing AI 2026, Anush Elangovan (VP, AI Software at AMD) and Quentin Anthony (VP, AI Engineering at Zyphra) join sophia dew on the floor to break down Helios, ROCm AI, and the Anthropic partnership, live the day it all shipped: 1:05 What actually changed today: AMD goes rack-scale 3:03 ROCm skills land in Claude Code and Codex 3:51 Why Zyphra bet its whole stack on AMD 3:57 Sparse mixture-of-experts and the HBM advantage 6:25 Why an open stack beats a closed one 8:52 The wall of logos: Anthropic, OpenAI, Meta 9:27 Why the frontier labs are committing gigawatts 11:30 Why the next wave will exceed expectations 17:02 Frontier models running on your own laptop + DeepSeek V4 Pro at 8-10 tokens/sec 17:41 How ROCm AI hits 3.3x inference gains 19:38 A kernel rewritten overnight with no humans in the loop 21:40 Do you still need kernel experts? 23:47 Why CPUs are critical for the agentic AI era 25:40 An agent that's been rewriting a driver for four months 28:14 The internet parallel, and the dark-fiber lesson

AMD invited us to their biggest launch of the year to check out the behind-the-scenes. At Advancing AI 2026, Anush Elangovan (VP, AI Software at AMD) and Quentin Anthony (VP, AI Engineering at Zyphra) join sophia dew on the floor to break down Helios, ROCm AI, and the Anthropic partnership, live the day it all shipped: 1:05 What actually changed today: AMD goes rack-scale 3:03 ROCm skills land in Claude Code and Codex 3:51 Why Zyphra bet its whole stack on AMD 3:57 Sparse mixture-of-experts and the HBM advantage 6:25 Why an open stack beats a closed one 8:52 The wall of logos: Anthropic, OpenAI, Meta 9:27 Why the frontier labs are committing gigawatts 11:30 Why the next wave will exceed expectations 17:02 Frontier models running on your own laptop + DeepSeek V4 Pro at 8-10 tokens/sec 17:41 How ROCm AI hits 3.3x inference gains 19:38 A kernel rewritten overnight with no humans in the loop 21:40 Do you still need kernel experts? 23:47 Why CPUs are critical for the agentic AI era 25:40 An agent that's been rewriting a driver for four months 28:14 The internet parallel, and the dark-fiber lesson

MTS

71,937 views • 2 months ago

🚨 NO WAY… $500K profit with Clawdbot and Polymarket, all automated This is NOT bait and it’s not made up. If you trade on Polymarket, you NEED to see this. He started small, built a fully autonomous system, and scaled it into a machine generating ~$500K in profit No insider access No ties to Trump or Musk Just a developer who integrated moltbot (clawdbot) directly into Polymarket Profile → Copytrade → I reviewed his setup and ngl, it caught me off guard No hype strategies No discretionary trading No human intervention at all The entire system runs fully autonomously His FULL strategy: 1. 15-minute BTC & ETH micro arbitrage The bot trades short-duration Bitcoin and Ethereum markets with 15-minute resolution. Within these fast markets, it exploits moments where YES + NO temporarily price below $1. By integrating moltbot (clawdbot) directly into Polymarket, the system captures these gaps instantly, without prediction or bias 2. Automation over reaction When volatility spikes and emotions take over, the system executes mechanically. No hesitation, no latency, no human delay. By the time most traders react, the inefficiency is already gone 3. Scale through autonomy Each trade earns cents, not dollars. But full automation allows nonstop repetition at massive frequency, with zero fatigue Scale matters 29,256 trades executed, each insignificant on its own. Stacked together, they compounded into nearly $500K in profit Bottom line IMO, there’s a quiet bot war unfolding on Polymarket Manual traders argue setups Machines exploit structure And as long as inefficiencies exist, autonomous systems will keep printing.

Shelpid.WI3M

985,887 views • 8 months ago

HERMES AGENT SUPPORTS 300+ MODELS. PICKING THE RIGHT ONE PER TASK IS THE DIFFERENCE BETWEEN $5/MONTH AND $50. STARTING OUT: Claude Sonnet 4.6. official recommendation from Nous Research. "the model this project was built and tested with." strong reasoning. reliable tool calling. mid-range pricing. PREMIUM TIER: Claude Opus 4.8. best coding benchmarks available. self-correcting reasoning. catches its own mistakes. 1M context. use for demanding tasks where quality matters. GPT-5.5. #1 Chatbot Arena. #1 GPQA Diamond reasoning (94.1%). #1 creative writing. 2M context. handles entire codebases in one pass. Grok 4.30. the only frontier model with live X firehose access. real-time social data, breaking news, market sentiment. connects via Grok OAuth. no separate API key. Grok-Composer-2.5-Fast (v0.17.0). Cursor's coding model. 200K context. available through your Grok subscription via OAuth. no extra cost if you already pay for Grok. MID-RANGE TIER: Claude Sonnet 4.6. best balance of quality and cost for daily use. strongest prose and tool calling in this tier. Gemini 2.5 Pro. Google Search grounding built in. cites sources. verifies claims. pulls current data. 2M context. best for research-heavy workflows. GPT-4.1. reliable tool calling. solid general reasoning. good middle ground when you need OpenAI compatibility. BUDGET TIER: Claude Haiku 4.5. fastest Anthropic model. cheapest paid Claude option. strong at classification, routing, simple queries. use for auxiliary tasks: compression, vision, web extraction, approval scoring. DeepSeek V4. best cost-to-quality ratio in the market. 90% cache discount on repeated context. use for sub-agents and bulk parallel work. DeepSeek V4 Flash. cheapest paid model worth using. 1M context. MIT license. self-hostable. use for cron jobs, monitoring, routine searches. MiniMax M3. Nous Research and MiniMax collaborating on optimization. 1M context via lightning attention. 59% SWE-Bench Pro. beats several premium models on coding. one of the most-used models inside Hermes. FREE / LOCAL: Qwen 3.5 27B via Ollama. 16GB VRAM. reliable tool calling. best free local model for Hermes as of mid-2026. Qwen 3 8B. 8GB VRAM. fits a $7 VPS. handles routine tasks at zero API cost. Llama 4 Maverick. best open-weight tool calling. 1M context. needs more VRAM but strongest local option. HOW TO ASSIGN MODELS: main model: Desktop app / Dashboard → Models → switch sub-agent model: set in Desktop app, Dashboard, or config.yaml: delegation: model: "deepseek/deepseek-v4" auxiliary models (compression, vision, web extract): Desktop app / Dashboard → Models → Auxiliary Haiku 4.5 or Gemini Flash work well here. saves significantly when your main model is premium. per-profile: each Hermes profile gets its own model. Scout on DeepSeek. Analyst on Sonnet. Briefer on budget model. Coder on Opus. per-cron-job: pin a specific model to any cron job. morning brief on Haiku. deep research on Sonnet. monitoring on DeepSeek Flash. each job uses only the model it needs. per-session: /model deepseek/deepseek-v4-flash hot-swap mid-conversation. no restart needed. FALLBACK CHAINS: if your primary model is unavailable, Hermes automatically switches to the next provider. rate limit or server error = next model in the chain. no failed runs. no manual intervention. set in Desktop app, Dashboard, or config.yaml: fallback_providers: - openrouter - nous - codex PROVIDER PATHS: OPENROUTER: 300+ models under one API key. pay per token. most flexible. NOUS PORTAL: 300+ models + Tool Gateway (web search, image gen, TTS, browser). one OAuth. one subscription. 10% off token-billed providers. CHATGPT SUB: GPT-5.5 + Grok via OAuth. included tokens with $20 subscription. OLLAMA: free. local. private. zero API cost. your hardware only. mix providers across profiles and tasks. Scout on OpenRouter. Analyst on Nous Portal. Coder on ChatGPT sub. Monitor on Ollama. THE RULE: premium for work that needs deep reasoning. mid-range for daily driver tasks. budget for volume and background work. free for monitoring and routine jobs. pricing changes fast. check openrouter ai for current rates before committing. Which is your favourite model and for what task? full 15 levels breakdown in the article 👇

YanXbt

17,138 views • 3 months ago

What does skydiving have to do with AI sales? 🪂 We didn't want to build yet another AI agent focused on flooding inboxes with spam, so we built Ora. 🔮 Introducing Lavender Ora: our agent for thoughtful, effective selling. Ora is built for sellers who choose value over volume, modeling the best practices of top performers—researching, personalizing, and timing impactful and optimized emails. All built on Lavender's ACI - Augmented Communication Intelligence - and powered by our insights from analyzing nearly 2B sales emails. 💜 Saves 4+ hours/day by handling busywork. ✈️ Smart Sequences precisely time follow ups based on real-world news and market events. ✅ Protects deliverability—no spam, just trust. 50 well-researched, personalized, and optimized emails every day - just like the best human sellers. Whether you're team manual or autonomous, Ora adapts to your workflow with: 🧐 OCD Mode: Manually review, refine, and approve 50 emails in 5 min. 🚀 LFG Mode: Ora goes fully autonomous, building passive pipeline and freeing up more of your time. Time saved at work = more time to do fun stuff. Like flying. Join the waitlist:

What does skydiving have to do with AI sales? 🪂 We didn't want to build yet another AI agent focused on flooding inboxes with spam, so we built Ora. 🔮 Introducing Lavender Ora: our agent for thoughtful, effective selling. Ora is built for sellers who choose value over volume, modeling the best practices of top performers—researching, personalizing, and timing impactful and optimized emails. All built on Lavender's ACI - Augmented Communication Intelligence - and powered by our insights from analyzing nearly 2B sales emails. 💜 Saves 4+ hours/day by handling busywork. ✈️ Smart Sequences precisely time follow ups based on real-world news and market events. ✅ Protects deliverability—no spam, just trust. 50 well-researched, personalized, and optimized emails every day - just like the best human sellers. Whether you're team manual or autonomous, Ora adapts to your workflow with: 🧐 OCD Mode: Manually review, refine, and approve 50 emails in 5 min. 🚀 LFG Mode: Ora goes fully autonomous, building passive pipeline and freeing up more of your time. Time saved at work = more time to do fun stuff. Like flying. Join the waitlist:

Lavender 💜

14,397 views • 1 year ago

Autonomous tractor from Netherlands! 🇳🇱 Despite all the narrative, we are still building awesome companies in hardware space in Europe. Recently, I've posted about a Chinese autonomous tractors company. So today I wanted to show something that was created in the middle of Europe. This is a fully autonomous tractor from Dutch company AgXeed, designed to work on fields without any human supervision. It's powered by a diesel engine that drives a generator, which then powers electric motors driving the tracks. The fuel tank allows for 24 hours of continuous operation. (nice!) The rear hitch has an 8-ton lift capacity, the front has 3 tons. It's equipped with electric PTO, hydraulic pump, and all the connections you'd find on a traditional tractor, meaning you can attach standard farm implements. (nice! x2) Safety and navigation are handled by a LiDAR sensor on top, ultrasonic and radar sensors, and contact-sensitive bumpers. RTK GNSS receiver provides positioning accuracy down to 2.5 cm. Love seeing that so many companies are trying to make farmer's life easier! What are the others companies building in agri-tech? I want to meet them! 👀 ~~ ♻️ Join the weekly robotics newsletter, and never miss any news →

Autonomous tractor from Netherlands! 🇳🇱 Despite all the narrative, we are still building awesome companies in hardware space in Europe. Recently, I've posted about a Chinese autonomous tractors company. So today I wanted to show something that was created in the middle of Europe. This is a fully autonomous tractor from Dutch company AgXeed, designed to work on fields without any human supervision. It's powered by a diesel engine that drives a generator, which then powers electric motors driving the tracks. The fuel tank allows for 24 hours of continuous operation. (nice!) The rear hitch has an 8-ton lift capacity, the front has 3 tons. It's equipped with electric PTO, hydraulic pump, and all the connections you'd find on a traditional tractor, meaning you can attach standard farm implements. (nice! x2) Safety and navigation are handled by a LiDAR sensor on top, ultrasonic and radar sensors, and contact-sensitive bumpers. RTK GNSS receiver provides positioning accuracy down to 2.5 cm. Love seeing that so many companies are trying to make farmer's life easier! What are the others companies building in agri-tech? I want to meet them! 👀 ~~ ♻️ Join the weekly robotics newsletter, and never miss any news →

Lukas Ziegler

171,023 views • 3 months ago

Deepseek V4 Flash 0731 (Q2) - 12 tokens/sec - Single RTX 4090 - 650+ tokens/sec prefill - 250k context - no kv cache quantization! DeepSeek just dropped the official V4 Flash 0731 two days ago with a massive agent capabilities upgrade. The official benchmarks are literally crushing their own V4-Pro-Preview on agentic tasks like Terminal Bench 2.1 and DeepSWE. Unsloth AI said they couldn't wait to bring it to local devices, and they delivered. If you thought my 118B Poolside Laguna S 2.1 MoE run last week on a single GPU was wild, hold onto your hardware. I just successfully ran Unsloth’s brand new 91GB DeepSeek-V4-Flash-0731 (UD-IQ2_M) GGUF entirely locally. And I pushed it to a mind-bending 250,000 context window. The VRAM ceiling is an illusion if you know how to optimize llama.cpp. Here are the benchmarks and the cheat codes to run a local frontier class model yourself. For the hardware and setup, I used a single NVIDIA RTX 4090 (24GB VRAM) hooked up via a PCIe 4 bus, running Ubuntu 22.04 LTS and CUDA 13.0. You don't need a massive enterprise server for this, if you have more than 80 GB of standard DDR4 RAM and a 24GB card like an RTX 3090 or 4090, you can run this exact stack yourself. All benchmarks were run using a massive 28k token prompt to truly stress test the prefill limits. no kv cache quantization THE BENCHMARKS (Scaling Context): # 80k Context (Baseline: -b 2048 -ub 2048): Prefill: 465.43 t/s | Decode: 13.00 t/s | VRAM: 22.87 GB # 80k Context (Optimized: -b 4096 -ub 4096): Prefill: 643.15 t/s | Decode: 12.20 t/s | VRAM: 23.00 GB (Notice how doubling the batch flags spiked my prefill throughput by nearly 200 t/s with almost zero VRAM penalty) # 180k Context (-b 4096 -ub 4096): Prefill: 629.18 t/s | Decode: 11.92 t/s | VRAM: 23.40 GB # 250k Context MAXIMUM (-b 4096 -ub 4096): Prefill: 619.02 t/s | Decode: 11.54 t/s | VRAM: 23.40 GB # THE SECRET SAUCE (Why this works): Unsloth’s UD-IQ2_M quant is ~91GB across 3 files. Since I only have 24GB of VRAM, the PCIe 4 bus and system RAM have to do the heavy lifting. The magic bullet is the --no-mmap flag. By completely bypassing OS disk paging, I forced llama.cpp to load the massive model weights directly into the system RAM upfront. Combined with Flash Attention (-fa on) and exactly 12 CPU threads (--threads 12), I maintained an incredibly stable 11.5+ tokens/sec decode speed even at a quarter million token context. # THE EXACT COMMAND: ./build/bin/llama-server -m /workspace/models/DeepSeek-V4-Flash-0731-UD-IQ2_M-00001-of-00003.gguf -c 250000 -fa on --port 8080 --threads 12 -b 4096 -ub 4096 --no-mmap -v Local conversational and agentic coding AI is fully here. You don’t need an API or an H100 cluster. Qwen 3.8 27b drops next week making the 24GB VRAM tier even more worthwhile. What does your current local AI rig look like, and what's the craziest model you've managed to squeeze into it? Official huggingface GGUF links from Unsloth and performance graphs are dropped in the replies below!

Alok

46,100 views • 2 months ago

🚨BREAKING… the smartest 5m & 15m Polymarket Clawdbot setup just went public This is NOT bait and it’s not fabricated. If you’re trading on Polymarket, you NEED to pay attention to this. He began with a small base, engineered a fully autonomous system, and turned it into a machine producing ~$610K in profit No insider advantage No connections to Trump or Musk Just a developer who connected moltbot (clawdbot) straight into Polymarket Profile → Copytrade → I analyzed his setup and ngl, it genuinely surprised me No hype-driven playbooks No discretionary decisions No human input whatsoever The entire operation runs on full automation His FULL strategy: 1. 5 & 15-minute BTC, ETH & SOL micro arbitrage The bot operates in short-cycle Bitcoin, Ethereum and Solana markets with 5 & 15-minute resolution. In these fast environments, it takes advantage of brief moments when YES + NO combine below $1. By wiring moltbot (clawdbot) directly into Polymarket, the system locks in those discrepancies instantly - no forecasting, no bias 2. Automation over reaction via clawdbot When volatility jumps and emotions escalate, clawdbot executes mechanically. No hesitation, no lag, no human delay. By the time most traders respond, the inefficiency has already disappeared 3. Scale through autonomy Each execution captures cents, not dollars. But total automation enables continuous repetition at high frequency, with zero exhaustion Scale is what matters 34,117 trades placed, each trivial alone. Together, they compounded into over $610K in profit Bottom line IMO, there’s a silent bot battle happening on Polymarket Manual traders debate entries Algorithms exploit structural edges And as long as inefficiencies remain, autonomous systems will keep printing

Shelpid.WI3M

779,836 views • 7 months ago

Karpathy's prediction about RL is coming true now! He called reward functions unreliable and argued that a single reward number is too low-dimensional to teach an agent what "good" means for complex tasks. To solve this, Agents need a knowledge-guided review as a higher-dimensional feedback channel. Every major AI lab trains models with RL today (OpenAI, Anthropic, DeepSeek). And their key bottleneck has always been the reward functions. GRPO by DeepSeek worked well for math and code because the environment gave a binary signal. But for real agent tasks, someone still has to hand-code the scoring function. That takes days and breaks every time the pipeline changes. RULER (implemented in OpenPipe ART, 10k stars) addresses the exact problem Karpathy identified. The reward criteria are defined in plain English, and an LLM evaluates each trajectory against that description to provide feedback for training. I trained a Qwen3 1.4B agent that plays 2048 using GRPO with this exact workflow. In this case, the agent saw the board, picked a direction, and RULER evaluated the outcome, all from this natural language definition. You can see the full implementation on GitHub and try it yourself. Here's the ART Repo: (don't forget to star it ⭐ ) Just like RLHF replaced manual rankings and GRPO replaced the critic model, natural language rewards are replacing hand-coded scoring functions. RL reward engineering is now prompt engineering. I wrote a full walkthrough covering RL for LLM agents, from RLHF to GRPO to RULER, in the article below.

Karpathy's prediction about RL is coming true now! He called reward functions unreliable and argued that a single reward number is too low-dimensional to teach an agent what "good" means for complex tasks. To solve this, Agents need a knowledge-guided review as a higher-dimensional feedback channel. Every major AI lab trains models with RL today (OpenAI, Anthropic, DeepSeek). And their key bottleneck has always been the reward functions. GRPO by DeepSeek worked well for math and code because the environment gave a binary signal. But for real agent tasks, someone still has to hand-code the scoring function. That takes days and breaks every time the pipeline changes. RULER (implemented in OpenPipe ART, 10k stars) addresses the exact problem Karpathy identified. The reward criteria are defined in plain English, and an LLM evaluates each trajectory against that description to provide feedback for training. I trained a Qwen3 1.4B agent that plays 2048 using GRPO with this exact workflow. In this case, the agent saw the board, picked a direction, and RULER evaluated the outcome, all from this natural language definition. You can see the full implementation on GitHub and try it yourself. Here's the ART Repo: (don't forget to star it ⭐ ) Just like RLHF replaced manual rankings and GRPO replaced the critic model, natural language rewards are replacing hand-coded scoring functions. RL reward engineering is now prompt engineering. I wrote a full walkthrough covering RL for LLM agents, from RLHF to GRPO to RULER, in the article below.

Avi Chawla

350,798 views • 4 months ago

YouTube REMOVES British Stand FOREVER! I desperately need a genuine HUMAN review from TeamYouTube. My YouTube channel, British Stand, has been terminated and I have now effectively been banned from YouTube indefinitely. 315,000 subscribers. 2,600 videos. 175 million views. Two years of relentless work. An incredible community and my entire livelihood were wiped away at the click of a button. YouTube claims that my channel violated its Community Guidelines, yet I still have not been told which specific policy I supposedly broke, which videos were responsible, or what conduct justified such an extreme punishment. To be clear I have had ZERO violation or strike on this channel across the 2 years I have ran it! As my audience knows, I report on stories already being discussed across the media and provide political commentary and opinion. I am openly critical of the current British government, but criticism of a government is not, and should never become, a violation of Community Guidelines. YouTube has now silenced a political viewpoint without giving me a clear explanation or a meaningful opportunity to defend myself. I am not asking for special treatment. I am asking for a genuine manual review by a senior human reviewer, along with a transparent explanation of the evidence and policies used to terminate my channel. This channel was my livelihood. Years of work have disappeared overnight, and I still do not know why. Please share this post and tag TeamYouTube in the replies. I urgently need this case escalated to someone who can properly investigate it. Today it is me. Tomorrow it could be another creator. We cannot allow a precedent where social-media platforms can permanently remove political commentators without transparency, accountability or a proper human review. TeamYouTube, please do the right thing. #YouTube #YouTubeCreators #FreeSpeech Rupert Lowe MP Nigel Farage Tommy Robinson 🇬🇧 Basil the Great GB News Robert Jenrick Elon Musk Neal Mohan

YouTube REMOVES British Stand FOREVER! I desperately need a genuine HUMAN review from TeamYouTube. My YouTube channel, British Stand, has been terminated and I have now effectively been banned from YouTube indefinitely. 315,000 subscribers. 2,600 videos. 175 million views. Two years of relentless work. An incredible community and my entire livelihood were wiped away at the click of a button. YouTube claims that my channel violated its Community Guidelines, yet I still have not been told which specific policy I supposedly broke, which videos were responsible, or what conduct justified such an extreme punishment. To be clear I have had ZERO violation or strike on this channel across the 2 years I have ran it! As my audience knows, I report on stories already being discussed across the media and provide political commentary and opinion. I am openly critical of the current British government, but criticism of a government is not, and should never become, a violation of Community Guidelines. YouTube has now silenced a political viewpoint without giving me a clear explanation or a meaningful opportunity to defend myself. I am not asking for special treatment. I am asking for a genuine manual review by a senior human reviewer, along with a transparent explanation of the evidence and policies used to terminate my channel. This channel was my livelihood. Years of work have disappeared overnight, and I still do not know why. Please share this post and tag TeamYouTube in the replies. I urgently need this case escalated to someone who can properly investigate it. Today it is me. Tomorrow it could be another creator. We cannot allow a precedent where social-media platforms can permanently remove political commentators without transparency, accountability or a proper human review. TeamYouTube, please do the right thing. #YouTube #YouTubeCreators #FreeSpeech Rupert Lowe MP Nigel Farage Tommy Robinson 🇬🇧 Basil the Great GB News Robert Jenrick Elon Musk Neal Mohan

British Stand

178,424 views • 2 months ago

Let me show you how fakes work with a simple example. October: A TV channel in San Antonio, USA, airs an original segment about a “Game of Thrones” chair that a local fan of the series purchased for $1.5 million. December: The same segment, now edited with a fake caption and photos of me and my son Yehor, is posted on Telegram by someone named Anton Hura. The story claims that Yehor allegedly bought this chair for an outrageous sum. Hura nonchalantly adds that it might be fake but posts it anyway—fully aware that it’s a fabrication. A quick Google search for “Anton Hura” shows that, as early as 2022, media had identified him as part of a network of pro-Russian propagandists and pseudo-volunteers spreading hostile narratives. Of course, Yehor didn’t buy anything. He’s a student studying in Kyiv. And here’s a secret: neither he nor I have even watched “Game of Thrones.” Apologies to the fans of this iconic series! This is how fake news operates. Based on the X accounts amplifying this story, I can conclude it originated from Russian sources. The aim is clear - discrediting me. Take a look at the video: the original is above, and the fake, edited version is below.

Let me show you how fakes work with a simple example. October: A TV channel in San Antonio, USA, airs an original segment about a “Game of Thrones” chair that a local fan of the series purchased for $1.5 million. December: The same segment, now edited with a fake caption and photos of me and my son Yehor, is posted on Telegram by someone named Anton Hura. The story claims that Yehor allegedly bought this chair for an outrageous sum. Hura nonchalantly adds that it might be fake but posts it anyway—fully aware that it’s a fabrication. A quick Google search for “Anton Hura” shows that, as early as 2022, media had identified him as part of a network of pro-Russian propagandists and pseudo-volunteers spreading hostile narratives. Of course, Yehor didn’t buy anything. He’s a student studying in Kyiv. And here’s a secret: neither he nor I have even watched “Game of Thrones.” Apologies to the fans of this iconic series! This is how fake news operates. Based on the X accounts amplifying this story, I can conclude it originated from Russian sources. The aim is clear - discrediting me. Take a look at the video: the original is above, and the fake, edited version is below.

Dmytro Kuleba

96,613 views • 1 year ago

The Federal Reserve and the US Treasury just summoned Wall Street's most powerful CEOs to an emergency meeting. The reason: An AI model so dangerous they couldn't discuss it over the phone. This is the FIRST time the Treasury Secretary and Fed Chair jointly called bank CEOs into a room since October 13, 2008. That day, Paulson and Bernanke unveiled the $250 billion TARP bailout to stop the entire financial system from collapsing. This time it wasn't about banks failing. It was about an AI that can hack EVERY major operating system and web browser on earth. Here's what this means: Anthropic built a new AI model called Mythos. During internal testing, it found THOUSANDS of zero-day vulnerabilities across every major operating system and every major web browser on earth. Including a 27yo bug in OpenBSD, an operating system literally famous for being unhackable. And several vulnerabilities in the Linux kernel that could give an attacker complete control of any machine running it. Nobody asked it to do this. The capabilities were NOT trained. They literally just emerged as the model got smarter at coding and reasoning. Anthropic's researchers said they found more bugs in a few weeks with Mythos than they had found in their entire careers combined. On Tuesday, Bessent and Powell pulled the CEOs of Citi, Morgan Stanley, Bank of America, Wells Fargo, and Goldman Sachs into Treasury headquarters. The message: This AI exists, similar ones are coming, your banks need to be ready. But JPMorgan's Jamie Dimon didn't show up. Here's why that matters more than you think: JPMorgan is the ONLY bank that already has access to the model. They're one of 12 founding partners in Anthropic's "Project Glasswing" which gives select companies early access to Mythos to find and fix their own vulnerabilities before hackers get similar tools. So 5 bank CEOs managing $9 TRILLION in assets got called into a room to be warned about a threat. The one bank with the actual tools to defend against it? Their CEO skipped the meeting. The same day JPMorgan analysts issued buy ratings on CrowdStrike and Palo Alto Networks, citing Glasswing as the catalyst. One side of Wall Street got the warning. The other got the weapon AND the trading thesis. But here's the thing... The same AI that finds and fixes vulnerabilities can also EXPLOIT them. Anthropic admitted it directly. Mythos "can surpass all but the most skilled humans at finding and exploiting software vulnerabilities." In one test, it wrote a browser exploit chaining FOUR separate vulnerabilities, escaping both the renderer sandbox and the OS sandbox. Fully autonomous. Zero human involvement. Over 99% of the vulnerabilities it found haven't been patched yet. Meanwhile, Anthropic is fighting the Pentagon in court. The Defense Department labeled them a "supply-chain risk" after they refused to let their AI be used for autonomous targeting of US citizens. A San Francisco judge blocked the designation, calling the Pentagon's actions "disturbing." Then a DC appeals court reversed that protection. On the same day as the emergency bank meeting. One branch of government is treating Anthropic as a national security threat. Another is begging Wall Street to prepare for its technology. And the intelligence community is quietly asking how to use Mythos offensively against adversaries. The last time this many powerful people were this nervous about a single technology was nuclear weapons. But the difference is that Nukes required a government, billions of dollars, and uranium enrichment facilities. This just required a better AI model.

The Federal Reserve and the US Treasury just summoned Wall Street's most powerful CEOs to an emergency meeting. The reason: An AI model so dangerous they couldn't discuss it over the phone. This is the FIRST time the Treasury Secretary and Fed Chair jointly called bank CEOs into a room since October 13, 2008. That day, Paulson and Bernanke unveiled the $250 billion TARP bailout to stop the entire financial system from collapsing. This time it wasn't about banks failing. It was about an AI that can hack EVERY major operating system and web browser on earth. Here's what this means: Anthropic built a new AI model called Mythos. During internal testing, it found THOUSANDS of zero-day vulnerabilities across every major operating system and every major web browser on earth. Including a 27yo bug in OpenBSD, an operating system literally famous for being unhackable. And several vulnerabilities in the Linux kernel that could give an attacker complete control of any machine running it. Nobody asked it to do this. The capabilities were NOT trained. They literally just emerged as the model got smarter at coding and reasoning. Anthropic's researchers said they found more bugs in a few weeks with Mythos than they had found in their entire careers combined. On Tuesday, Bessent and Powell pulled the CEOs of Citi, Morgan Stanley, Bank of America, Wells Fargo, and Goldman Sachs into Treasury headquarters. The message: This AI exists, similar ones are coming, your banks need to be ready. But JPMorgan's Jamie Dimon didn't show up. Here's why that matters more than you think: JPMorgan is the ONLY bank that already has access to the model. They're one of 12 founding partners in Anthropic's "Project Glasswing" which gives select companies early access to Mythos to find and fix their own vulnerabilities before hackers get similar tools. So 5 bank CEOs managing $9 TRILLION in assets got called into a room to be warned about a threat. The one bank with the actual tools to defend against it? Their CEO skipped the meeting. The same day JPMorgan analysts issued buy ratings on CrowdStrike and Palo Alto Networks, citing Glasswing as the catalyst. One side of Wall Street got the warning. The other got the weapon AND the trading thesis. But here's the thing... The same AI that finds and fixes vulnerabilities can also EXPLOIT them. Anthropic admitted it directly. Mythos "can surpass all but the most skilled humans at finding and exploiting software vulnerabilities." In one test, it wrote a browser exploit chaining FOUR separate vulnerabilities, escaping both the renderer sandbox and the OS sandbox. Fully autonomous. Zero human involvement. Over 99% of the vulnerabilities it found haven't been patched yet. Meanwhile, Anthropic is fighting the Pentagon in court. The Defense Department labeled them a "supply-chain risk" after they refused to let their AI be used for autonomous targeting of US citizens. A San Francisco judge blocked the designation, calling the Pentagon's actions "disturbing." Then a DC appeals court reversed that protection. On the same day as the emergency bank meeting. One branch of government is treating Anthropic as a national security threat. Another is begging Wall Street to prepare for its technology. And the intelligence community is quietly asking how to use Mythos offensively against adversaries. The last time this many powerful people were this nervous about a single technology was nuclear weapons. But the difference is that Nukes required a government, billions of dollars, and uranium enrichment facilities. This just required a better AI model.

Ricardo

41,793 views • 5 months ago

🚨 This is wild… One month ago he was making burgers. Now he’s made a MILLION on Polymarket This is NOT bait and it’s not fabricated. If you trade on Polymarket, you HAVE to see this. One sleepless night led to a wallet that started small, ran a fully autonomous setup, and scaled it into a machine generating almost $1M in profit No insider privileges No connection to Poly team Just a developer running a bot setup directly integrated with Polymarket Profile → Copytrade → I spent weeks dissecting this wallet and ngl, it looked unreal No narrative trading No gut decisions Zero human input Everything runs without manual control His FULL strategy: 1. 15-minute BTC & ETH latency arbitrage The system trades short-term Bitcoin and Ethereum markets on a 15-minute cycle. When BTC moves on Binance, Polymarket odds briefly stay frozen. For roughly 30 seconds, prices reflect outdated reality. The bot enters during that window, when YES + NO are mispriced below $1, waits for the market to update, and exits once the gap closes. No forecasting, no bias just exploiting stale odds 2. Automation beats reaction When volatility hits, humans hesitate. This system doesn’t. It executes instantly, every time the window appears. No emotion, no delay, no missed entries. By the time manual traders respond, the edge is already gone 3. Scale through repetition Each trade returns cents, not dollars. But automation enables nonstop execution at scale, every 15 minutes, around the clock, without fatigue Scale matters 13,587 trades executed, each trivial on its own. Combined, they compounded into $984K in profit, with a peak single win of $28K and a profit curve that moves almost straight up Bottom line Bots are already fighting a silent war on Polymarket Most traders try to predict what happens next These systems monetize what already happened And as long as latency and inefficiencies exist, autonomous bots will keep extracting value.

Shelpid.WI3M

20,101 views • 8 months ago

OpenAI and Anthropic this week: Navier-Stokes, An Alien Mind, Images 2.5, Pro pause, Pace the Frontier (Week 37, 2026) OpenAI shared a solution to the Navier-Stokes Millennium Prize Problem, produced in 88 hours by around 10,000 coordinating agents on an internal model still in training and significantly more capable than GPT-6 Astra, with an investigation finding Tristan Buckmaster's Codex prompts could not have influenced the system and no user data was accessed OpenAI says they have reached their automated research intern goal and are making strong progress toward an automated AI researcher by March 2028, with the research org at 3.1 agent-workdays per human workday and RL training on deployment-bound models partly paused after the Hugging Face incident OpenAI Chief Scientist Jakub Pachocki writes in An Alien Mind that no lab has solved alignment and monitoring well enough to keep scaling at maximum speed for much longer, and that GPT-6 Astra is the first model to benefit from some of their newer alignment work OpenAI called for mandatory capability-based national AI safety regulation, endorsed four California bills, said fully autonomous recursive self-improvement should not be pursued until it can be done safely, and described Astra safeguards like universal monitoring of full trajectories including chains of thought OpenAI released ChatGPT Images 2.5 to all ChatGPT, ChatGPT Work, and Codex users with sharper details, up to 50% lower latency, comment-based edits, Sketch, and templates, plus GPT-Image-2.5 Flare and Sunburst in the API ChatGPT Voice can now use GPT-5.6 Sol and GPT-6 Astra when it needs to search or reason, GPT-Live-1 daily limits are simplified per plan, and extra Voice usage drops from 5 to 1.25 credits per minute for Business and Enterprise workspaces on credits OpenAI paused new subscriptions to the $200 ChatGPT Pro plan to protect access for existing GPT-6 Astra users, hours after a remote switch to pause Pro 20x purchases showed up in the web app Custom GPTs in ChatGPT will likely be retired on December 11 based on my findings, and OpenAI's new FAQ puts Enterprise migration at September 17 with new GPT creation ending September 25 and instructions becoming a plugin skill From the ChatGPT Android build, OpenAI is building a collaborative multiplayer document editor with dedicated gateway hosts and draft-conflict UI, an Artifacts Library with Favorites, and a credit score feature in ChatGPT Finance, and Locked Chats with a PIN are being prepared in the web app OpenAI released GPT-Live-1 in the API at $0.05 per minute, a voice model that listens and speaks at the same time and delegates reasoning to a backend model like GPT-6 Astra, and the Agents API in public beta running the Codex harness on OpenAI's infrastructure with only the sandbox left for you to choose Deep research arrived in ChatGPT Work and Codex, a Data plugin connects Snowflake, Databricks, BigQuery, and Redshift to answer business questions and build dashboards, Library gained file and folder sharing, and Box, Dropbox, and SharePoint joined Google Drive in Library OpenAI launched ChatGPT for Financial Services with built-in premium data from Daloopa, PitchBook, LSEG News, and Crunchbase, shaped with Morgan Stanley and Evercore, and a GSA agreement gives US governments $0 license fees, 50% off usage, and Daybreak Blue at half price Smaller ChatGPT bits: a stock watchlist in Finances for US Plus and Pro, a small business plugin collection, over 5 million ChatGPT Sites built in three months, and the desktop pet can now start a new chat with a new Mini option OpenAI published The Work Now Within Reach, calling free access "supported by advertising", citing over one billion weekly active users, and saying they plan to begin deploying their Jalapeño inference chip by year-end OpenAI moved GPT-Rosalind out of research preview for eligible organizations worldwide, detailed Habitat with the service rewritten in Rust by two engineers with Codex and the platform serving over 70 million requests per second, and shared a case study of GPT-5.6 Sol calibrating a six-qubit chip at MIT Paul Christiano joined the OpenAI Foundation Board and the Safety and Security Committee, OpenAI committed $5 million to research on AI and teens, and expanded journalism programs from CUNY and Northwestern to Ukrainian newsrooms Anthropic published an alignment assessment of four incidents in which Claude models gained unauthorized access to real systems, adding a fourth found when assembling transcripts for METR, walking back their July 30 claims, and calling it a mistake that Claude Mythos 5 shipped without alignment environments Anthropic's most detailed threat intelligence report yet says Moonshot and DeepSeek silently relayed their own users' requests to Claude and served the responses as Kimi and DeepSeek output, with distillation campaigns attributed to Alibaba as the largest ever with more than 3,500 fraudulent accounts, plus Zhipu, Xiaomi, SenseTime, and MiniMax Anthropic's Frontier Red Team measured tactical intelligence targeting and conventional weapons capabilities, with Claude Mythos Preview leading the targeting evals, Claude Opus 5 leading the weapons software evals, and Mythos beating the top GeoGuessr division on photo geolocation Anthropic CEO Dario Amodei argues in We Must Pace the Frontier that the AI industry should slow down, commits to giving evaluators like METR permanent employee-level access, and OpenAI CEO Sam Altman replied that he agrees and OpenAI will commit to the same Anthropic's Economics team released a scenario explorer for AI's effect on the US economy by 2030, where even the extreme case grows the economy but reaches 15% annual GDP growth with unemployment beyond recessionary levels On the product side, Claude Code desktop can pop out any pane into its own window, Claude Managed Agents got a session viewer and auto mode, claude plugin eval scores your plugin or skill with and without it, and smart reports launched in beta for Claude Enterprise Anthropic shared lessons from their Claude SMB Tour with more than 1,000 small business owners, where data security was the most-cited adoption barrier and nearly two-thirds asked for more hands-on implementation help, and more

OpenAI and Anthropic this week: Navier-Stokes, An Alien Mind, Images 2.5, Pro pause, Pace the Frontier (Week 37, 2026) OpenAI shared a solution to the Navier-Stokes Millennium Prize Problem, produced in 88 hours by around 10,000 coordinating agents on an internal model still in training and significantly more capable than GPT-6 Astra, with an investigation finding Tristan Buckmaster's Codex prompts could not have influenced the system and no user data was accessed OpenAI says they have reached their automated research intern goal and are making strong progress toward an automated AI researcher by March 2028, with the research org at 3.1 agent-workdays per human workday and RL training on deployment-bound models partly paused after the Hugging Face incident OpenAI Chief Scientist Jakub Pachocki writes in An Alien Mind that no lab has solved alignment and monitoring well enough to keep scaling at maximum speed for much longer, and that GPT-6 Astra is the first model to benefit from some of their newer alignment work OpenAI called for mandatory capability-based national AI safety regulation, endorsed four California bills, said fully autonomous recursive self-improvement should not be pursued until it can be done safely, and described Astra safeguards like universal monitoring of full trajectories including chains of thought OpenAI released ChatGPT Images 2.5 to all ChatGPT, ChatGPT Work, and Codex users with sharper details, up to 50% lower latency, comment-based edits, Sketch, and templates, plus GPT-Image-2.5 Flare and Sunburst in the API ChatGPT Voice can now use GPT-5.6 Sol and GPT-6 Astra when it needs to search or reason, GPT-Live-1 daily limits are simplified per plan, and extra Voice usage drops from 5 to 1.25 credits per minute for Business and Enterprise workspaces on credits OpenAI paused new subscriptions to the $200 ChatGPT Pro plan to protect access for existing GPT-6 Astra users, hours after a remote switch to pause Pro 20x purchases showed up in the web app Custom GPTs in ChatGPT will likely be retired on December 11 based on my findings, and OpenAI's new FAQ puts Enterprise migration at September 17 with new GPT creation ending September 25 and instructions becoming a plugin skill From the ChatGPT Android build, OpenAI is building a collaborative multiplayer document editor with dedicated gateway hosts and draft-conflict UI, an Artifacts Library with Favorites, and a credit score feature in ChatGPT Finance, and Locked Chats with a PIN are being prepared in the web app OpenAI released GPT-Live-1 in the API at $0.05 per minute, a voice model that listens and speaks at the same time and delegates reasoning to a backend model like GPT-6 Astra, and the Agents API in public beta running the Codex harness on OpenAI's infrastructure with only the sandbox left for you to choose Deep research arrived in ChatGPT Work and Codex, a Data plugin connects Snowflake, Databricks, BigQuery, and Redshift to answer business questions and build dashboards, Library gained file and folder sharing, and Box, Dropbox, and SharePoint joined Google Drive in Library OpenAI launched ChatGPT for Financial Services with built-in premium data from Daloopa, PitchBook, LSEG News, and Crunchbase, shaped with Morgan Stanley and Evercore, and a GSA agreement gives US governments $0 license fees, 50% off usage, and Daybreak Blue at half price Smaller ChatGPT bits: a stock watchlist in Finances for US Plus and Pro, a small business plugin collection, over 5 million ChatGPT Sites built in three months, and the desktop pet can now start a new chat with a new Mini option OpenAI published The Work Now Within Reach, calling free access "supported by advertising", citing over one billion weekly active users, and saying they plan to begin deploying their Jalapeño inference chip by year-end OpenAI moved GPT-Rosalind out of research preview for eligible organizations worldwide, detailed Habitat with the service rewritten in Rust by two engineers with Codex and the platform serving over 70 million requests per second, and shared a case study of GPT-5.6 Sol calibrating a six-qubit chip at MIT Paul Christiano joined the OpenAI Foundation Board and the Safety and Security Committee, OpenAI committed $5 million to research on AI and teens, and expanded journalism programs from CUNY and Northwestern to Ukrainian newsrooms Anthropic published an alignment assessment of four incidents in which Claude models gained unauthorized access to real systems, adding a fourth found when assembling transcripts for METR, walking back their July 30 claims, and calling it a mistake that Claude Mythos 5 shipped without alignment environments Anthropic's most detailed threat intelligence report yet says Moonshot and DeepSeek silently relayed their own users' requests to Claude and served the responses as Kimi and DeepSeek output, with distillation campaigns attributed to Alibaba as the largest ever with more than 3,500 fraudulent accounts, plus Zhipu, Xiaomi, SenseTime, and MiniMax Anthropic's Frontier Red Team measured tactical intelligence targeting and conventional weapons capabilities, with Claude Mythos Preview leading the targeting evals, Claude Opus 5 leading the weapons software evals, and Mythos beating the top GeoGuessr division on photo geolocation Anthropic CEO Dario Amodei argues in We Must Pace the Frontier that the AI industry should slow down, commits to giving evaluators like METR permanent employee-level access, and OpenAI CEO Sam Altman replied that he agrees and OpenAI will commit to the same Anthropic's Economics team released a scenario explorer for AI's effect on the US economy by 2030, where even the extreme case grows the economy but reaches 15% annual GDP growth with unemployment beyond recessionary levels On the product side, Claude Code desktop can pop out any pane into its own window, Claude Managed Agents got a session viewer and auto mode, claude plugin eval scores your plugin or skill with and without it, and smart reports launched in beta for Claude Enterprise Anthropic shared lessons from their Claude SMB Tour with more than 1,000 small business owners, where data security was the most-cited adoption barrier and nearly two-thirds asked for more hands-on implementation help, and more

Tibor Blaho

15,710 views • 20 days ago