Matthew John Brady

Founder & CEO of @VolleySolutions; Instructor @LeedsBiz. Serial entrepreneur with 4 successful exits. Edu: MBA @KelloggSchool, BS @TechPurdue.

[Matthew John Brady @_mattbrady]

The Volley app for Students (https://t.co/i1cMYNRQtD) is now available, providing easy tools for tough decisions. Access the app in any browser or download to any device. It's free for academic and personal use.
https://t.co/nHGCtUFDnA

[Matthew John Brady @_mattbrady]

@victormarx Amen

[Alex Prompter @alex_prompter]

🚨 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

[Greg Laurie @greglaurie]

According to YouVersion, the most engaged-with verse of 2025 was Isaiah 41:10:

“Do not fear, for I am with you;
do not be dismayed, for I am your God.
I will strengthen you and help you;
I will uphold you with My righteous right hand.”

That’s God’s promise—and it means we don’t have to be afraid as we step into this new year.

We may not know what the future holds,
but we know who holds the future.

[HD28 Jeffco @JeffcoHD28]

Become a Precinct Committee Person.
Great resource for how and why!
https://t.co/u43MlLd0I9

[Victor Marx @victormarx]

I’m Victor Marx, and I’m running to be the next Governor of Colorado

It’s time we lead with courage, protect families, restore values, and prioritize people over politics

JOIN THE MOVEMENT👇🏼
https://t.co/PVnEZK4HBP

[Larry P. Arnn @DrLarryArnn]

I delivered these remarks at Charlie Kirk’s memorial service.

[ATLANTA TECH BLOGS @ATLTechBlogs]

New post from AssureSign: Nintex and CU Boulder’s Leeds School of Business collaborate to launch groundbreaking application development program for student innovation #atlantatechnews #technology #startups https://t.co/nmoh4TRLOK

[Sticker Mule @stickermule]

WE ARE GIVING AWAY A CYBERTRUCK

ACTUALLY A CYBERBEAST

REPOST TO ENTER

FOLLOW SO WE CAN DM YOU

NO PURCHASE NECESSARY. TERMS: https://t.co/P9OJtf9RCw

[Anthony Pompliano 🌪 @APompliano]

America Has Become A Society Of Gamblers

You can see this everywhere you look. Zero-day options are dominating public markets. Meme coins are bought and sold like 1980’s boiler room penny stocks. Major sport leagues are integrating sports betting into every corner of their operations.

Risk taking might as well be the 28th amendment to the US Constitution at this point.

This is what happens when a nation’s currency is being destroyed. People continue to fall behind financially and the wealth inequality gap widens. Rather than seeing a path to financial security, the average person loses hope.

There is no path to a brighter future.

Once hope is lost, the gambler’s mindset takes over. So what if they lose a little money on a 100:1 odds bet? The alternative is a life of debt and financial pressure.

Gambling is the best of two bad choices — at least the individual feels a degree of agency when selecting which bet to make. This is insane behavior when viewed through a logical lens, but it makes complete sense when you view it through human psychology.

Gambling gives people hope in a broken economy.

You may not think that the US economy is broken, but the average American would disagree. It is now less expensive to rent than purchase in all 50 major metros. Citizens owe more than $1.1 trillion on credit card balances. More than 43 million Americans have federal student loan debt, with the average balance being more than $37,000.

The US dollar has lost 25% of its purchase power since January 2020. For every $1 someone had 4 years ago, they can only buy $0.75 worth of goods today.

Given these facts, gambling doesn’t seem so illogical.

There are only two paths to fixing the problem. Either the United States figures out how to stop the accelerated debasement of the currency, which would allow citizens to gain restored confidence in the protection of their hard-earned economic value, or citizens will have to resist the urge to gamble and instead seek out alternative assets that help them outperform the inflation they are experiencing.

It is hard to tell someone to buy the S&P 500 when they feel like they have to find 10x investment opportunities to stay ahead.

Singles and doubles have created many millionaires. Home run swings lead to many strike outs. But this data doesn’t matter when people have lost hope. They need the light at the end of the tunnel, which is being shined brightest today by zero-day options, meme coins, and sports betting.

Many of us may not agree with it. We must acknowledge what is happening though.

The future damage created by this development is hard to comprehend. An entire generation is growing up in a different regime and it is unlikely they can be coaxed back into value investing, dollar-cost averaging, and other timeless investing principles.

YOLO and FOMO rule the day. Such a sad reality.

[David Sacks @DavidSacks]

Best way to listen to the Milei speech in English, rendered using HeyGen, in his own voice.

[Luke Sophinos @lukesophinos]

It's not very often that you see a company take ~10 years to get to ~$4M in revenue, and then absolutely explode to $250M+.

A vision ahead of its time, that ultimately caught up to the rest of the world.

This is the story of Procore:

[Matthew John Brady @_mattbrady]

I enjoyed working with another @Vistage key group. I can suggest the right #PowerofVistage membership to help you advance your career and your impact -- https://t.co/zhV64Lc55A. Book a #DecideBetter workshop to equip your team in #DecisionScience -- https://t.co/7i7uMu4jqr

[Matthew John Brady @_mattbrady]

I recently spoke with a group of @Vistage executives, with a packed room for the #DecideBetter workshop. The impact that #DecisionScience will have for them is profound. Thanks to our hosts at @Food4LarimerCo. Please support your local food bank. #PowerofVistage #Leadership

[Matthew John Brady @_mattbrady]

I recently spoke to a dynamic group via the @Vistage CEO network. It was a robust conversation about #DecisionScience, #ExecutiveLeadership, and opportunities for #Improvement using the @VolleySolutions collaborative decision optimization platform. #DecideBetter #PowerofVistage

[ETHDenver 🏔🦬🦄 @EthereumDenver]

HELLO FELLOW BUFFICORNS!

THE WAIT IS FINALLY OVER

APPLY TODAY FOR ETHDenver 2023: #TheYearOfTheSpork at https://t.co/dJBKkIVmBN!

We can't wait to see you next year at the National Western Complex!

[Denver Westword @denverwestword]

This #Denver #climbing gym owner needs help keeping the lights on, because when kids are 'problem-solving in this space, they're problem-solving on their tests.'
https://t.co/YMt3Ak7oky

[Matthew John Brady @_mattbrady]

Thanks to the @Vistage Chief Executive and Key Executive groups that I spoke to this week. #DecisionScience improves processes and outcomes. It's an emerging field that can increase engagement, compliance, and competitive advantage. #DecideBetter

[Rodney Smith Jr @iamrodneysmith]

Hello CO ! Great to be back . We will be in the Denver area delivering gifts in the morning . We are looking for a family in need of some Christmas cheer here. Know any ?