Thomas Bloom

Royal Society University Research Fellow at the University of Manchester. Mathematician and owner of He/him/his.

[Thomas Bloom @thomasfbloom]

@nekoma_shu "We" includes all of humanity! (whatever the career stage, and whether mathematician or not)

Everyone has different parts of the new they'd like to take advantage of, and faces different risks, so this is a difficult problem.

[Thomas Bloom @thomasfbloom]

Coincidentally, this morning I watched Queen Elizabeth's first televised Christmas message in 1957. This part struck me as still relevant:

"That it's possible for some of you to see me today is just another example of the speed at which things are
changing all around us...

1/5

[Thomas Bloom @thomasfbloom]

@AcerFur @BlackHC I've heard various versions of what happened (and is continuing to happen now) from different people 'inside'.

A strange situation; will make a fascinating retrospective one day.

(As will the inside story of e.g. OpenAI during 2026, if that ever gets to come out.)

[Thomas Bloom @thomasfbloom]

@GregHBurnham Maybe the thing they didn't like about Old England was our overly confrontational attitude to foliage.

[Thomas Bloom @thomasfbloom]

@avery_yen @GregHBurnham Why do Americans and the Japanese have seasonal plant viewing as enshrined cultural traditions, yet the British do not?

[Thomas Bloom @thomasfbloom]

@wtgowers @nihilunbounded Yes, this would also be valuable, and I hope some people will do this.

(Just not me - personally there's more than enough interesting stuff to explore in the 'classical' non-PDE approach.)

[Thomas Bloom @thomasfbloom]

The Queen also went on to give religion/morality as her examples of these ageless ideals which were being discarded.

I don't think the same applies here.

My general point is "how do we take advantage of the new without losing the best of the old" is the right question.

5/5

[Thomas Bloom @thomasfbloom]

(This was because I have been watching The Crown and got curious about that period of British history.

It is not usual for British people to start their day by watching old speeches by the Queen.)

4/5

[Thomas Bloom @thomasfbloom]

@nihilunbounded Actually, me too. I'm maybe more excited to do this now, since (a) we know the big conjectures are true, so our intuition are not wrong, and (b) since there exists a proof maybe Big AI will leave it alone, letting humans explore and develop ideas organically.

[Thomas Bloom @thomasfbloom]

@ColinZwanziger @wtgowers Yes, fair enough, would be better to say Lean as an example. (I guess they focused on Lean as the most popular, hence the least likely to contain bugs etc. If AI formalises a result in a little used language without much documentation or human expertise, doesn't help confidence.)

[Thomas Bloom @thomasfbloom]

@nihilunbounded Then think about why that's true if all the lines exactly line up. Then think about what can change if we start to angle the lines slightly away from each other (but only O(1/n) away).

[Thomas Bloom @thomasfbloom]

@nihilunbounded If you're new to this kind of statement/argument, my tip is to start by setting S=1, ignore the epsilon stuff, and divide both sides by the product of N_j^{1/n-1}. Now the RHS is O(1) and the LHS is the integral of the product of expectations.

[Thomas Bloom @thomasfbloom]

@ColinZwanziger @wtgowers Because it favours Lean over other possible formalisation languages you mean?

[Thomas Bloom @thomasfbloom]

@josebrox @rperezmarco Ah yes, I agree that I implicitly was, of course, thinking of a time variable. So the question may be better phrased as "what problems do current frontier AI models find difficult?"

[Thomas Bloom @thomasfbloom]

@josebrox @rperezmarco I don't understand this argument. The same would apply to almost any use of data or scientific research, since everything changes all the time.

That doesn't stop a snapshot of data from one moment (especially a big watershed moment like this) being important/useful.

[Thomas Bloom @thomasfbloom]

@ivanich22893 @wtgowers To clarify, that screenshot is from a conversation with Sebastien Bubeck, who is not part of AGMAI, so it doesn't make sense to describe his comment as a position 'abandoned' by AGMAI.