Tim

bsky:

[Ed Zitron @edzitron]

Uber’s COO has said that it’s getting “harder to justify” its AI costs because there was no way to show a link between AI spend and any meaningful increase in useful features. This is the first time I’ve seen a company say this directly.

https://t.co/xUhZvtpwah

[Grady Booch @Grady_Booch]

I think that @DarioAmodei does not understand software engineering and that he is working feverishly to pump up the valuation of his company in anticipation of its forthcoming IPO.

[unusual_whales @unusual_whales]

BREAKING: Just 20 minutes before Trump's announcement that the Strait of Hormuz was open, massive trades hit the market.

Investors sold a combined 7,990 lots of Brent crude futures, ​a $760 million bet that oil would go down.

These orders were much larger than anything else at the time.

The traders made huge gains.

Unusual.

[Nav Toor @heynavtoor]

🚨SHOCKING: Apple just proved that AI models cannot do math. Not advanced math. Grade school math. The kind a 10-year-old solves.

And the way they proved it is devastating.

Apple researchers took the most popular math benchmark in AI — GSM8K, a set of grade-school math problems — and made one change. They swapped the numbers. Same problem. Same logic. Same steps. Different numbers.

Every model's performance dropped. Every single one. 25 state-of-the-art models tested.

But that wasn't the real experiment.

The real experiment broke everything.

They added one sentence to a math problem. One sentence that is completely irrelevant to the answer. It has nothing to do with the math. A human would read it and ignore it instantly.

Here's the actual example from the paper:

"Oliver picks 44 kiwis on Friday. Then he picks 58 kiwis on Saturday. On Sunday, he picks double the number of kiwis he did on Friday, but five of them were a bit smaller than average. How many kiwis does Oliver have?"

The correct answer is 190. The size of the kiwis has nothing to do with the count.

A 10-year-old would ignore "five of them were a bit smaller" because it's obviously irrelevant. It doesn't change how many kiwis there are.

But o1-mini, OpenAI's reasoning model, subtracted 5. It got 185.

Llama did the same thing. Subtracted 5. Got 185.

They didn't reason through the problem. They saw the number 5, saw a sentence that sounded like it mattered, and blindly turned it into a subtraction.

The models do not understand what subtraction means. They see a pattern that looks like subtraction and apply it. That is all.

Apple tested this across all models. They call the dataset "GSM-NoOp" — as in, the added clause is a no-operation. It does nothing. It changes nothing.

The results are catastrophic.

Phi-3-mini dropped over 65%. More than half of its "math ability" vanished from one irrelevant sentence.

GPT-4o dropped from 94.9% to 63.1%.

o1-mini dropped from 94.5% to 66.0%.

o1-preview, OpenAI's most advanced reasoning model at the time, dropped from 92.7% to 77.4%.

Even giving the models 8 examples of the exact same question beforehand, with the correct solution shown each time, barely helped. The models still fell for the irrelevant clause.

This means it's not a prompting problem. It's not a context problem. It's structural.

The Apple researchers also found that models convert words into math operations without understanding what those words mean. They see the word "discount" and multiply. They see a number near the word "smaller" and subtract. Regardless of whether it makes any sense.

The paper's exact words: "current LLMs are not capable of genuine logical reasoning; instead, they attempt to replicate the reasoning steps observed in their training data."

And: "LLMs likely perform a form of probabilistic pattern-matching and searching to find closest seen data during training without proper understanding of concepts."

They also tested what happens when you increase the number of steps in a problem. Performance didn't just decrease. The rate of decrease accelerated. Adding two extra clauses to a problem dropped Gemma2-9b from 84.4% to 41.8%. Phi-3.5-mini from 87.6% to 44.8%. The more thinking required, the more the models collapse.

A real reasoner would slow down and work through it. These models don't slow down. They pattern-match. And when the pattern becomes complex enough, they crash.

This paper was published at ICLR 2025, one of the most prestigious AI conferences in the world.

You are using AI to help you make financial decisions. To check legal documents. To solve problems at work. To help your children with homework. And Apple just proved that the AI is not thinking about any of it. It is pattern matching. And the moment something unexpected shows up in your question, it breaks. It does not tell you it broke. It just quietly gives you the wrong answer with full confidence.

[Grady Booch @Grady_Booch]

I like Claude; I use Claude daily. But what Anthropic is doing is only a small part of the activities of a software engineering.

[Jesse McCormick @jessecmccormick]

A Federal Court judge has upheld the government's decision that far-right media outlet Rebel News doesn't qualify for journalism tax credits because it doesn't produce enough original content.
https://t.co/SbCcN00UOc

[Scott Robertson @sarobertson_]

HÉBERT: "The last thing we need are for opposition members to suddenly freelance and say we're gonna renegotiate the relationship [with the US] and come home to say 'Donald Trump told me he likes us.'"

[Larry Short, CPA @Larry_Short]

If anybody knows of a family that needs a Christmas tree in the St. John’s area, I have a perfectly good one that I can deliver in the next couple of hours. Let me know right away. This one already has the lights on it and everything works. It’s just that I found a second one in our shed.

[Scott Fitzgerald @scottyfitzNL]

Hey folks. I'm excited to announce I'll be running for Councillor At Large in the upcoming St. John's municipal election. Check out my website and let me know what I can work on for you. https://t.co/w9BRXGfITh

[Transportation and Infrastructure NL @TI_GovNL]

🧹 Spring cleaning is coming for the Outer Ring Road! On Sunday, May 25, a litter cleanup and highway closure is planned for the ORR from Kenmount Road to the Logy Bay Road interchange from 7:00 a.m. to 7:00 p.m.

Details: https://t.co/eNZMr6zU8N

#GovNL #nltraffic

[Steve Coombs 🇨🇦 @Stephen_Coombs]

Check out Kneaded https://t.co/ZHOSptnpJj

[Tim @tb_jam]

@NBA Boston is one of the worst teams to watch. Live and die by the three

[unusual_whales @unusual_whales]

The S&P500 experienced its biggest first 100 day loss for a President since 1973, per WSJ:

[Curtis Traverse 🏒🏳️‍🌈 ❌ 🇨🇦 @z56cjt]

Kevin page from @IFSD_IFPD did a great explanation of this tonight on @PnPCBC Conservatives are basically including magic revenue in their numbers to make the deficits look better. Educate yourself before voting if this is important to you.

[chantal hébert @ChantalHbert]

+100

[Tim @tb_jam]

Well nice to know what this guy thinks of folks in NL. @NTVNewsNL still hosts his segment no matter how many people he degrades.

[Dan McTeague @GasPriceWizard]

Is it the pogey or govt jobs?

[Rick Mercer @rickmercer]

Tilting at windmills Newfoundland style. Personally my Canada includes a bit of Land & Sea. #Nlpoli #DiscoverNL #CdnPoli #NLvotes #NL #SkipperAndCompany #ListofLights

[chantal hébert @ChantalHbert]

Édifiant: Poilievre Stop in Fish Plant Smells | Canadian Journalist https://t.co/XgloCo44Ws

[PressProgress @pressprogress]

NEW: Conservative candidate used a secret Signal group chat with Freedom Convoy leaders, right-wing media and far-right influencers

Andrew Lawton was a member of the “clandestine” group chat used to coordinate convoy messaging on social media

https://t.co/rWP60klPoG #cdnpoli

[AA1973 | 🇺🇦 🇪🇺 🇮🇱 | @Andreas_Adam]

Listen up!
Scott Lucas on the most illiterate speeches and economically most damaging actions he ever heard of.
👊🐧🔥

Who's Scott Lucas?
👇
https://t.co/kvkOIz4WKz