Working4Canada

Trying My Best

[Working4Canada @Working4Canada]

@CAPE_ACEP Excellent work! Keep it up!

[CAPE - ACEP @CAPE_ACEP]

We look forward to MP Fanjoy and others pushing back and ultimately saving Canadians the $40 billion dollars over ten years it had projected when the Liberals promised to turn more office spaces into much-needed housing instead.

[Working4Canada @Working4Canada]

AI in the form of LLMs are scientifically proven to be inherently unreliable.

[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

[Olya Povlatsky 🇵🇱🇨🇦🇺🇦 @GongshowRefugee]

@PierrePoilievre Bruce Fanjoy has launched a petition to the H of C for all federal party leaders to obtain and maintain top security clearance (which PP is staunchly refusing).

Open for signing until April 7th. 25,000 signatures in just 3 days!
Please sign and repost.

https://t.co/pK7CF7T4zf

[Working4Canada @Working4Canada]

@tallmikeholley @grok @psac_afpc Hah, even Grok is telling you you're wrong Mike. Try thinking for yourself.

[Working4Canada @Working4Canada]

@Roman_Baber @MarkJCarney Where is the honorable member for BattleRiver-Crowfoot in all this? Why hasn't there been an official statement supporting a more efficient public service?

[Working4Canada @Working4Canada]

@psac_afpc @AlexSilasPSAC @SharonDPSAC Will you be doing the same and meeting with @brucefanjoy ?

[CAPE - ACEP @CAPE_ACEP]

We’re meeting today with MP Bruce Fanjoy and look forward to hearing more about how we can push back against the new RTO policy together.

[Working4Canada @Working4Canada]

About time the government got called out on that!

[PSAC-AFPC @psac_afpc]

$6 billion. That’s what remote work could save taxpayers.
Carney’s government knows it. Their own research shows it.

So why force workers back into overcrowded offices? Stop the waste.
Join the fight to protect remote work. https://t.co/BPcCsH8zXo

[Working4Canada @Working4Canada]

@tallmikeholley @brucefanjoy @grok Can't think for yourself eh Mike?

[Working4Canada @Working4Canada]

@BlokeQuebecois @brucefanjoy Ingratitude? If you think the PS is a sweet deal, fill out an application to join.

[Working4Canada @Working4Canada]

@mcintyre_sg @brucefanjoy You're flagrantly wrong Susan. Post some facts with sources, not your opinions.

[Working4Canada @Working4Canada]

Thank you @brucefanjoy for standing up for Canadians who want a more efficient public service!

[Bruce Fanjoy 🍁 @brucefanjoy]

Please see my statement regarding the Government’s updated Return-to-Office policy.

[Bruce Fanjoy 🍁 @brucefanjoy]

Please see my statement regarding the Government’s updated Return-to-Office policy.

[PSAC-AFPC @psac_afpc]

Destaffing light stations on the Coast of British Columbia is not the answer. There is no substitute for their physical presence as our eyes and ears on the coast. Sign this letter to tell Minister Lebouthillier to review her decision. https://t.co/SprZq12Cb4

[Volodymyr Zelenskyy / Володимир Зеленський @ZelenskyyUa]

Dear @JustinTrudeau, I appreciate your kind words and unwavering support on Ukraine’s Independence Day. The connection between Ukraine and Canada runs deep, rooted in the bonds between our people and the shared values we hold dear. Together, we stand strong and united in our defense of freedom, democracy, and rules-based international order. 🇺🇦🇨🇦

[Mark Gerretsen 🇨🇦 🇺🇦 @MarkGerretsen]

If y'all want to know why Conservatives routinely vote against Ukraine, just read the comments from PP's own base on his tweet. 👇

[Teamsters Canada @TeamstersCanada]

‼️The CIRB's decision sets a dangerous precedent: big companies can now pause operations briefly, and the government will break unions. Workers' rights are significantly diminished. The Teamsters will comply, but our fight is just beginning.

Learn more 👇
https://t.co/p4yWY32FrV

[Jagmeet Singh @theJagmeetSingh]

Former Conservative Minister just called the Liberal Governments crushing of Rail Workers’ rights  “remarkable for the employer.”

Yeah, and garbage for workers.

Both the Conservatives and Liberals will always sellout workers to CEOs.

[Working4Canada @Working4Canada]

@TeamstersCanada For all Canadians, I hope you win your appeal! The government should not be in the business of union busting. They should be supporting workers' rights!

[CAPE - ACEP @CAPE_ACEP]

CAPE President Nathan Prier defending public sector workers in today's @OttawaCitizen: Sutcliffe plays blame game instead of solving transit problems https://t.co/fsD8Idq82H #CndPoli #CanLab