Brady Long
RT @thisdudelikesAI: I've been copy + pasting ChatGPT responses into emails like some kind of digital assembly line worker...

Meanwhile this AI is in iMessage catching contract errors and booking flights autonomously.

I'm so behind... but Lindy's about to change that: https://t.co/ueXITRiX5v https://t.co/JYk610Rqn7

Introducing Lindy Assistant, the ultimate AI assistant.
It talks with you through iMessage, connects to 100s of apps, helps you with your meetings and emails, and proactively finds ways to save you time all day.
Check out some examples of ways Lindy assistant helps below. https://t.co/DY10QQSSfn - Flo Crivello
tweet

God of Prompt
RT @p_song1: Thank you for sharing our work on Large Language Model Reasoning Failures!

📄 Paper: https://t.co/qKTwAsIBz9
📚 Repo: https://t.co/KOSB6811LH

🚨 Holy shit… Stanford just published the most uncomfortable paper on LLM reasoning I’ve read in a long time.

This isn’t a flashy new model or a leaderboard win. It’s a systematic teardown of how and why large language models keep failing at reasoning even when benchmarks say they’re doing great.

The paper does one very smart thing upfront: it introduces a clean taxonomy instead of more anecdotes. The authors split reasoning into non-embodied and embodied.

Non-embodied reasoning is what most benchmarks test and it’s further divided into informal reasoning (intuition, social judgment, commonsense heuristics) and formal reasoning (logic, math, code, symbolic manipulation).

Embodied reasoning is where models must reason about the physical world, space, causality, and action under real constraints.

Across all three, the same failure patterns keep showing up.

> First are fundamental failures baked into current architectures. Models generate answers that look coherent but collapse under light logical pressure. They shortcut, pattern-match, or hallucinate steps instead of executing a consistent reasoning process.

> Second are application-specific failures. A model that looks strong on math benchmarks can quietly fall apart in scientific reasoning, planning, or multi-step decision making. Performance does not transfer nearly as well as leaderboards imply.

> Third are robustness failures. Tiny changes in wording, ordering, or context can flip an answer entirely. The reasoning wasn’t stable to begin with; it just happened to work for that phrasing.

One of the most disturbing findings is how often models produce unfaithful reasoning. They give the correct final answer while providing explanations that are logically wrong, incomplete, or fabricated.

This is worse than being wrong, because it trains users to trust explanations that don’t correspond to the actual decision process.

Embodied reasoning is where things really fall apart. LLMs systematically fail at physical commonsense, spatial reasoning, and basic physics because they have no grounded experience.

Even in text-only settings, as soon as a task implicitly depends on real-world dynamics, failures become predictable and repeatable.

The authors don’t just criticize. They outline mitigation paths: inference-time scaling, analogical memory, external verification, and evaluations that deliberately inject known failure cases instead of optimizing for leaderboard performance.

But they’re very clear that none of these are silver bullets yet.

The takeaway isn’t that LLMs can’t reason.

It’s more uncomfortable than that.

LLMs reason just enough to sound convincing, but not enough to be reliable.

And unless we start measuring how models fail not just how often they succeed we’ll keep deploying systems that pass benchmarks, fail silently in production, and explain themselves with total confidence while doing the wrong thing.

That’s the real warning shot in this paper.

Paper: Large Language Model Reasoning Failures - God of Prompt
tweet

Moon Dev
oh codex is the same as claude code

such an interesting phenomenon rn

ai is being subsidized by investors

but at some point claude has to cut you off

and when they do

you have to switch to the other guy

to see all this ai is relatively the same

bless to be a dev rn
tweet

Brady Long
RT @bigaiguy: Everyone's hyped about Clawdbot

But I’m not smart enough to:

1. Buy a Mac Mini (I refuse)
2. Security configs (?)
3. Managing API bills (??)
4. Explain to IT why I need (Not today)

I just want the assistant so I use https://t.co/9XNXEnsJM5 https://t.co/wH21OhrMpC

Introducing Lindy Assistant, the ultimate AI assistant.
It talks with you through iMessage, connects to 100s of apps, helps you with your meetings and emails, and proactively finds ways to save you time all day.
Check out some examples of ways Lindy assistant helps below. https://t.co/DY10QQSSfn - Flo Crivello
tweet

Dimitry Nakhla | Babylon Capital®
RT @BarrySchwartzBW: I think the market is wrong here on some infotech names. Ratings by trusted authorities becomes much more valuable in an AI driven world. Owning proprietary data sets becomes much more valuable as AI cannot disrupt it.
tweet

Fiscal.ai
Robinhood Q4 Results

Total Revenues +27%
Adj. EPS +22%
Total Assets +68%
Funded Customers +7%
Average Revenue per User +16%
Gold Subscribers +58%

$HOOD: -6.7% AH https://t.co/eo1PjeHk67
tweet

Clark Square Capital
RT @ClarkSquareCap: There are about 400 ideas to sort through, including divestments, M&A, IPOs, rights offerings, and spin-offs. So there's something for everyone. Be sure to check it out.

Feedback welcome!

Sharing a new project: the Special Situations Digest.

Check out the (free) link below. https://t.co/NT0wb21Sxl - Clark Square Capital
tweet

Fiscal.ai
Looks like Prediction Markets are quite the hit at Robinhood.

$HOOD https://t.co/ScMfW9vwvF
tweet