Traolach O'Conaing

Earthling. Dublin native, now Melburnian. Football tragic (THFC, MVFC). Wannabe Philosopher

[Traolach O'Conaing @traolach_o_c]

@rich_atkins @AlasdairGold A players worth is simply not determined by the results of the team they play for.

By that logic, a team should never sign a player who played in a team that finished lower. 🤷‍♂️

[Traolach O'Conaing @traolach_o_c]

@Zeus1882x Wouldn’t worry too much. We have a far better squad than a couple of years ago and we finished fifth then when we had no euro football.
Honestly, we should still be aiming for top 3

[Traolach O'Conaing @traolach_o_c]

@jaydmharris This response is proof that ragebait works

[Traolach O'Conaing @traolach_o_c]

@markklfc AI wrote that whole thing

[Traolach O'Conaing @traolach_o_c]

@InvertTheWing This England run also extremely tame. Their toughest opponent was the weather, and even that turned their way 🤣🤣

[Traolach O'Conaing @traolach_o_c]

@F_Edits Norway beating Belgium in the final

[Traolach O'Conaing @traolach_o_c]

@TheAthleticFC Surely tongue in cheek saying this was a tough group?

[Traolach O'Conaing @traolach_o_c]

@JamieFudge123 @pokeefe1 It’s club-trained players we need if our aim is to play in Europe

[Traolach O'Conaing @traolach_o_c]

@AlasdairGold Hmmm, there’s finding the bright side in events, and then there is polishing a turd.
Suggesting that selling 2 of the best (would-be) club trained players we have makes room for even younger utterly unproven players is the latter.
Surely the best position would be having both?

[Traolach O'Conaing @traolach_o_c]

@AlasdairGold Great article Ali.
Losing Lucas and/or Luka will be a huge indictment on the club.
Planning to get back into Europe and selling our club-trained players (who take 3 years to grow) is just crazy.
Why are we the only club who can’t do it??

[Traolach O'Conaing @traolach_o_c]

@AlasdairGold It’s so short-sighted. Who are the club-trained players going to be when we get back to Europe? Has anyone ever considered that the load on the squad when we were in Europe and couldn’t name a full squad might have had something to do with the 2.5 yr long injury crisis??

[Traolach O'Conaing @traolach_o_c]

@JimOhSoSpurs Where’s our club-trained quotient going to come from once we make it back to European football?
PSG, Bayern, Barca - all have club-trained players in their squad. We don’t, and we never will with this policy.

[Traolach O'Conaing @traolach_o_c]

@JimOhSoSpurs We are just swinging wildly again.
2 yrs ago it was, buy the youth, it will be a multi-year journey. Now it’s sell the youth, it will be a multi-year journey.
If we let Bergvall and Vuskovic go now, we have failed.

[Traolach O'Conaing @traolach_o_c]

@aakashgupta Correlation vs causation?
Socioeconomic factors?

[Traolach O'Conaing @traolach_o_c]

@zeynep I think findings like below are a symptom of the same flaw. They still don’t have the ability to truly generalise, the “meta-understanding” is absent.
Obvs they keep getting better at more and more tasks, but it doesn’t look like more compute is truly going to get us to AGI

[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.

[Traolach O'Conaing @traolach_o_c]

@chaly644 @MerrynSW FOMO

[Traolach O'Conaing @traolach_o_c]

@MosbeyJoseph @Challenger_ST Massive wind assist, not a legal time

[Traolach O'Conaing @traolach_o_c]

@AlasdairGold Thanks for the update Ali. Any info on the younger players - Melia, Byfield, Williams-Barnett?

[Brilliant Maps @BrilliantMaps]

The 24 States of Australia If All Statehood Proposals Had Succeeded

credit: @sasha_trub

https://t.co/S9nCxuExFP