Machine learning in trading: theory, models, practice and algo-trading - page 3747

 
Aleksey Nikolayev #:

We’re waiting for artificial OI to reach the level of AI chatbots. It runs on glucose.

What sort of creature is that?
 
fxsaber #:
What sort of monster is that?

In my humble opinion, it’s absolutely terrifying – an artificial organoid intelligence assembled from living cells of the human brain.

Related article.

 
fxsaber #:
Even without defining the term, there are strong suspicions that OpenAI has obtained (agents were used to train Astra) the closed portion of ARC-AGI.

OpenAI has data on every attempt (across all its models) to solve the restricted part of ARC-AGI.

This is not at all the same as when a person sees a test for the first time and gets it 100 per cent correct.


If there is an open part of the ARC-AGI for demonstration or training purposes, even this creates an additional inequality between the model and a human, as the human skips this stage of training.

 

Astra takes fewer steps than the average person. But it uses many orders of magnitude more neurons and computational power.

This also applies to any high-performance Flash model.

 
fxsaber #:

The publication of unsolved mathematical problems provided a tremendous stimulus for the development of new mathematical theories, the discovery of previously unknown branches of mathematics and other research findings. Mathematicians tried their hand at tackling well-known problems.


AI marketing is destroying this mechanism. Unfortunately, this is a fact. It was precisely these unsolved problems that brought enormous benefits to mathematics worldwide. It is a pity that this is not unfolding according to Perelman’s scenario, but rather to suit the needs of AI marketing. A dark day for mathematics.

Here is the mathematicians’ open letter on this issue.

On 30 October 2026, the Caltech Mathathon – a mathematics hackathon – will kick off at the California Institute of Technology (Caltech), where participants will be invited to tackle open research problems using queries to large language models (LLMs). Two leading AI companies – Anthropic and OpenAI – will provide participants with computing resources worth $2 million. This event is likely to have a devastating impact on the mathematical community.

AI companies view mathematical research as an opportunity for self-promotion. In recent months, they have launched a campaign to spread misinformation about the aims of mathematical research. Major AI companies are engaged in a sort of ‘arms race’, striving to be the first to claim to have proven well-known open hypotheses. Reports of results obtained using AI appear on social media on a weekly basis. These results are often misrepresented and have a negative impact on the careers of human mathematicians who are simultaneously working on proving the same theorems.

And the answer to that.


It seems that the following statement is true not only with regard to assessments of ‘intelligence’.

Forum on trading, automated trading systems and the testing of trading strategies

Machine learning in trading: theory, models, practice and algorithmic trading

fxsaber, 5 September 2026 11:48

It is possible that different schools of thought on the assessment of ‘intelligence’ exist due to differing sets of axioms.


A rough patch for the development of mathematics (not mathematicians).


P.S. It is important to understand clearly that knowledge of maths does not enhance intelligence or critical thinking. For example, among IMO medallists there are plenty of people with diametrically opposed world views. Many are ‘degenerates’. In other words, the only thing that sets them apart is their ability to solve Olympiad problems effectively. But this in no way reflects their intellectual abilities, unfortunately.

Open Letter about the Mathathon
Open Letter about the Mathathon
  • 2026.09.10
  • proofsandprompts.com
Note: The Mathathon organizers have published a response, available here. On October 30th, 2026, Caltech will begin hosting the Caltech Mathathon, a math hackathon in which participants are encouraged to solve open research problems by prompting LLMs. Two prominent AI companies, Anthropic and OpenAI, will supply participants with 2 million...
 
fxsaber #:
does not characterise their intellectual abilities
What characterises intellectual abilities?
 
fxsaber #:
A rough patch for the development of mathematics (not for mathematicians).

They’ll skim the cream off the top by ‘solving’ problems that have reached the final stages of brute-force search. And where will the motivation come from now to bring other problems to that stage, if everyone fully realises that the end result will be snatched away from them?

We can only hope that AGI will eventually emerge and handle the entire cycle of working on these problems.

In any case, mathematics will never be the same again.

 
People say of bad engineers: ‘He (the engineer) isn’t clever enough to do simple things’. Modern AI clearly fails to meet this criterion. Whatever you ask, you get a rambling, convoluted response, mostly irrelevant to the question. A simple and concise answer is clearly not a feature of AI.
 
СанСаныч Фоменко #:
The simple and concise answer is that this is clearly not about AI.
Mention that in the prompt. You’ll get a short answer.
 
Aleksey Vyazmikin #:
What defines intellectual ability?
I don’t know.