LogoThread Easy
  • 発見
  • スレッド作成
LogoThread Easy

Twitter スレッドの万能パートナー

© 2025 Thread Easy All Rights Reserved.

探索

Newest first — browse tweet threads

Keep on to blur preview images; turn off to show them clearly

People under-estimate how much harder it would be to align agents that learn purely from their own experience to our values.  As much as LLM Alignment has become a cottage industry of sorts, it's child play in comparison! #NeurIPS2025

People under-estimate how much harder it would be to align agents that learn purely from their own experience to our values. As much as LLM Alignment has become a cottage industry of sorts, it's child play in comparison! #NeurIPS2025

AI researcher & teacher @SCAI_ASU. Former President of @RealAAAI; Chair of @AAAS Sec T. Here to tweach #AI. YouTube Ch: https://t.co/4beUPOmf6y Bsky: rao2z

avatar for Subbarao Kambhampati (కంభంపాటి సుబ్బారావు)
Subbarao Kambhampati (కంభంపాటి సుబ్బారావు)
Wed Dec 03 19:35:51
Nearly 30k people registered for NeurIPS (64% increase from last year).

This is what a single row of posters looks like. There are hundreds of these across six sessions.

Exhibit hall is similarly packed. The most popular booth may be the @Tesla Optimus + neural simulation!

Nearly 30k people registered for NeurIPS (64% increase from last year). This is what a single row of posters looks like. There are hundreds of these across six sessions. Exhibit hall is similarly packed. The most popular booth may be the @Tesla Optimus + neural simulation!

Partner @a16z AI 🤖 and twin to @omooretweets | Investor in @elevenlabsio, @krea_ai, @bfl_ml, @hedra_labs, @wabi, @WaveFormsAI, @ViggleAI, & more

avatar for Justine Moore
Justine Moore
Wed Dec 03 19:32:29
Excited to continue to help support insitro and CEO/founder @DaphneKoller in their pioneering applications of AI to bio and health!

Excited to continue to help support insitro and CEO/founder @DaphneKoller in their pioneering applications of AI to bio and health!

Co-Founder at @VZVCfund. Investing in companies solving humanity’s biggest problems, led by founders hell bent on bending the curve.

avatar for Vijay Pande
Vijay Pande
Wed Dec 03 19:31:16
RT @clairevo: Sorry babes, you were all wrong. 

I put Gemini 3, Claude Opus-4.5, and GPT-5.1 Codex head to head in a one shot design battl…

RT @clairevo: Sorry babes, you were all wrong. I put Gemini 3, Claude Opus-4.5, and GPT-5.1 Codex head to head in a one shot design battl…

Deeply researched product, growth, and career advice

avatar for Lenny Rachitsky
Lenny Rachitsky
Wed Dec 03 19:31:09
it’s easy to blame regulation as the sole culprit for why we can’t build things fast or cheaply anymore, but that argument misses a deeper truth about eroding industrial capacity.

take the vogtle nuclear reactors. they weren’t derailed  by the NRC alone, but by failures much closer to the ground: repeated construction-management turnover, an untrained and unstable workforce, a supply chain that effectively had to be rebuilt from scratch, and the decision to start pouring concrete and assembling modules before the designs were even finished. crews made bad welds, improperly poured concrete, delivered out-of-spec modules, and even failed to procure long-lead components on time. the NRC certainly didn’t accelerate timelines, but it would be a mistake to assign all of the blame to the regulator when the core breakdown occurred inside the engineering, procurement, and construction process itself.

the real bottleneck is at the EPC level — the firms responsible for coordinating the engineering, procurement, and construction of america’s largest capital projects. this is where our national capacity has eroded. we no longer have the industrial muscle once embodied by builders like henry kaiser, and without simultaneously rebuilding this capability, no amount of deregulation will be enough.

the good news is that software and AI now give us the opportunity to rethink this sector. by modernizing the EPC stack — from planning and design to procurement and on-site execution — we can finally deliver complex projects at the speed and quality the moment demands.

@unltdindustries is leading that transformation, and we’re proud to support them.

it’s easy to blame regulation as the sole culprit for why we can’t build things fast or cheaply anymore, but that argument misses a deeper truth about eroding industrial capacity. take the vogtle nuclear reactors. they weren’t derailed by the NRC alone, but by failures much closer to the ground: repeated construction-management turnover, an untrained and unstable workforce, a supply chain that effectively had to be rebuilt from scratch, and the decision to start pouring concrete and assembling modules before the designs were even finished. crews made bad welds, improperly poured concrete, delivered out-of-spec modules, and even failed to procure long-lead components on time. the NRC certainly didn’t accelerate timelines, but it would be a mistake to assign all of the blame to the regulator when the core breakdown occurred inside the engineering, procurement, and construction process itself. the real bottleneck is at the EPC level — the firms responsible for coordinating the engineering, procurement, and construction of america’s largest capital projects. this is where our national capacity has eroded. we no longer have the industrial muscle once embodied by builders like henry kaiser, and without simultaneously rebuilding this capability, no amount of deregulation will be enough. the good news is that software and AI now give us the opportunity to rethink this sector. by modernizing the EPC stack — from planning and design to procurement and on-site execution — we can finally deliver complex projects at the speed and quality the moment demands. @unltdindustries is leading that transformation, and we’re proud to support them.

investing @a16z // curating https://t.co/ssslqn6eo7

avatar for Ryan McEntush
Ryan McEntush
Wed Dec 03 19:31:05
RT @AhmedHemaz: DSPy integration with mlflow for prompt versioning, evaluation

RT @AhmedHemaz: DSPy integration with mlflow for prompt versioning, evaluation

Asst professor @MIT EECS & CSAIL (@nlp_mit). Author of https://t.co/VgyLxl0oa1 and https://t.co/ZZaSzaRaZ7 (@DSPyOSS). Prev: CS PhD @StanfordNLP. Research @Databricks.

avatar for Omar Khattab
Omar Khattab
Wed Dec 03 19:28:29
  • Previous
  • 1
  • More pages
  • 1677
  • 1678
  • 1679
  • More pages
  • 5634
  • Next