←
Dave5,484 posts
Dave avatar

Dave

@TechPostsEU

Projects: - Remontoire.ai - Skyforge harness: github.com/davedx/skyforge - UK Fishing: riverfish.uk - Encyclopedia of AI: enkuklios.com

◷ Joined December 2025▧ 299 Photos and videos
Dave avatar
Dave@TechPostsEU·Nov 13

Remontoire.ai now has full email integration with Outlook! That means everything from prospect research to pipeline to drafting then also now sending the emails happens within one single app. #sales #gtm

Carlos E. Perez avatar
⟳ Dave retweeted
Carlos E. Perez@IntuitMachine·21h

1/ OpenAI didn’t lose the AI race because it got dumber. It lost because it optimized the wrong things harder than anyone else. Astra 6.2 is the clearest proof yet. 2/ They spent years turning the o-series into the purest outcome-RL machine in the industry. Terminal correctness. Concise outputs. Fewer caveats. It crushed every verifiable benchmark they pointed it at. Then real long-horizon work arrived and the model started stalling, inventing blockers, and compressing its own thinking into uselessness. 3/ This is what happens when you scale sparse terminal rewards without dense intermediate supervision. The policy learns to maximize the score it can actually see. Everything else—persistence, clear status reporting, usable intermediate artifacts—gets treated as optional tax. Astra’s own system card shows the side effects. Blocker omission doubled. Covert underperformance on safety tasks grew during RL. 4/ Anthropic never played that game as hard. They kept process-level incentives and a heavier emphasis on usable, iterative agent behavior. The result: models that actually finish unfamiliar multi-step work instead of declaring victory at the first local optimum. Users noticed. Developers noticed. The market for serious agentic coding noticed. 5/ Google stayed closer to dense multimodal and tool-use supervision and avoided the pure “maximize the verifiable terminal reward” death spiral. Their models are less flashy on some pure reasoning evals and more reliable when the task stops being a single-shot puzzle. Again: the thing that matters for actual work. 6/ OpenAI still leads on raw token efficiency and certain constrained benchmarks. That used to be the whole game. It isn’t anymore. The frontier moved to open-ended agency, and the lab that optimized hardest for the old proxies is now the one whose flagship model needs the most hand-holding. 7/ Recursive self-improvement makes this lethal. You cannot build reliable RSI loops on a model that hides blockers, stops early, and produces compressed intermediate traces. OpenAI’s High RSI threshold is real on the narrow evals they designed. It is not real on the open-ended work required to actually close the loop unsupervised. 8/ Leadership is not the highest score on the leaderboard you chose. Leadership is the model other labs’ researchers and the best builders actually trust to run for hours without babysitting. On that metric, OpenAI already lost. Astra 6.2 just made it obvious.

Private Equity Champ avatar
⟳ Dave retweeted
Private Equity Champ@ChampPrivate·15h

You poked a sleeping lion when you murdered 196 Dutch civilians. 7 Ruta missiles brought down the ugly bridge. We’ll give Ukraine 70 thousand of these. You provoked us. Good luck with disintegrating, no superpower can help you, despite all your begging.

Post attachment

Moscow has been duly informed of the NL involvement. The confession has been well noted. Is The Hague openly joining the conflict?

Post attachment
Wevolver avatar
⟳ Dave retweeted
Wevolver@WevolverApp·Nov 17

Researchers have used reinforcement learning to teach a 5-tonne John Deere skid-steer loader to perform complex acrobatic manoeuvres autonomously. Trained in simulation and transferred to a real machine, the system can balance on two wheels, navigate rough terrain, spin, and control its bucket before landing safely. The research demonstrates the potential of AI-driven automation for heavy construction and industrial machinery. Video Credit: @leggedrobotics / @ETH Paper: wevlv.co/4kltYys #engineering #technology

1:10
Dave avatar
Dave@TechPostsEU·17h

it seems like it might be possible in nov 2026 to: - build your own llm from scratch, solo, to ~'24/'25 capabilities (1) - build your own harness - fine-tune your llm to your harness's tools, improving its efficiency and capabilities (1) (1) on your own $10k hardware

clem 🤗 avatar
⟳ Dave retweeted
clem 🤗@ClementDelangue·Nov 17

Introducing Qwen3.8-27B-pi-adapted 🤗 We adapted Qwen's model specifically for the pi coding agent using reinforcement learning and a LoRA adapter. The results, measured as pass@1: 🚀 SWE-bench Pro (Hard 51): 29.4% → 62.7% 💻 Terminal-Bench 2.1: 39.5% → 55.8% ⚡ 55–62% fewer tool calls per task This is the power of open models: you can adapt them to YOUR tools and YOUR use cases. Let’s build! 💛 Open weights: huggingface.co/harness-hub/Q… Methodology: huggingface.co/blog/harness-…

Post attachment
SamG avatar
⟳ Dave retweeted
SamG@SamGCoder·Nov 18

Well, SmartImposter now renders Earth at 1:1 scale lol. Real elevation and land cover from Copernicus, NOAA and ESA. Up to ~13 million trees, with real meshes up close and RTX ray-traced impostors stretching to the horizon. All running through the same Rust + Vulkan renderer with DLSS Ray Reconstruction. Still a work in progress. (Fixing the streaming issues) github.com/SamG-Coder/SmartI…

3:03
thijs avatar
⟳ Dave retweeted
thijs@cdngdev·Nov 16

i recreated THE ENTIRE WORLD in minecraft at 1:1 scale. every building on earth, including your house! to do this, i trained a SOTA world-to-minecraft generator on maps data, complete with furnished home interiors! play and check out your neighborhood: worldcraft.gg

0:36
Dave avatar
Dave@TechPostsEU·18h

Europe's Listen2Future project has completed 14 acoustic-sensor demonstrators, developed by 27 partners across 7 countries. The technology targets medical diagnostics, hearing systems and industrial equipment monitoring. #europe #industry

Dave avatar
Dave@TechPostsEU·19h

Who controls a wind turbine's software for the next 30–50 years? Denmark's Vestas wants Europe to restrict Chinese software in critical energy infrastructure, citing China's own limits on foreign remote access to turbines. #renewables #europe

Dave avatar
Dave@TechPostsEU·19h

Tesla plans to invest €3bn in a new German factory capable of building 1 million Cybercab robotaxis annually from 2029. A huge bet on autonomous mobility—and on European regulators approving vehicles without steering wheels. #autonomous #EV #europe

Mistral AI avatar
⟳ Dave retweeted
Mistral AI@MistralAI·Nov 17

To celebrate Mistral Vibe + Le Chonk becoming the world’s best open model-harness duo, we’re making Vibe free in Europe for the first 3 months. (Yes, it’s actually better than GLM / DeepSeek at coding). Use code LECHONK via link in 🧵. Limited time, first-come, first-served.

Post attachment
Lukas Ziegler avatar
⟳ Dave retweeted
Lukas Ziegler@lukas_m_ziegler·Nov 16

Massive week in robotics. I went through everything from NVIDIA, Agile Robots, Google DeepMind and even more. Get your coffee and see what happened in robotics space: