AI

This account is verified. Learn more

@DeepLearn007

Imtiaz Adam CS #AI Postgrad |#Strategy #MachineLearning RSI Evolutionary Neurosymbolic |#RL #Agentic | #LLM Liberal | MBA alum @morganstanley @LBS @Columbia_Biz

linkedin.com/in/imtiaz-adam… Joined September 2012

109K Following 135K Followers 67K Posts

Latest posts 5

AI @DeepLearn007 1d

I expect @OpenAI to launch Sol 6.2 and Astra 6.2 during the @WebSummit as it’s a very significant global tech event.

Delighted to be attending and happy to meet those also there or if even if not there let’s build a vibrant AI startup community and keep the spirit of the Web Summit.

Web Summit Verified account @WebSummit 2d

Opening Night is tomorrow! 🇵🇹

Please be aware that Centre Stage has limited capacity and is likely to fill up early.

See you there!

View quoted post on X

AI @DeepLearn007 2d

This is why you have to be careful about some social media posts as factual

Binkg Verified account @dummerspast39 3d

BREAKING: Thibault Sottiaux (@thsottiaux), better known as Tibo, has left OpenAI.

He has removed all mentions of OpenAI from his bio.

View quoted post on X

AI @DeepLearn007 6d

The key point is to not cede management or executive decisions to AI. You cannot fire an AI for a bad decision such as one that could result in material financial
losses, death and destruction.

But AI can work alongside humans for better outcomes and result in enhanced decision making through collaboration.

Sam Altman Verified account @sama 6d

People should always be in control of the future.

OpenAI is against the notion that it would be acceptable for people to be relegated to a role of doing things that “AI systems think makes sense to reward in humans”.

People should not cede control of the future to superintelligence, and that requirement should ultimately constrain AI progress.
View quoted post on X

AI @DeepLearn007 Nov 2

MIT researchers using Astra 6 to enhance Qwen 3.5 0.8 in relation to ANN weights

Rikiya Takehi @rikiyatakehi Nov 2

We asked GPT-6 Astra to edit Qwen3.5-0.8B weights to improve math.
It directly changed only 0.016% of weights, fixing 0.8B's reasoning loops & improving math up to 6x and even instruction following 3x, without gradients or training.

We call this Weight-Space Programming: frontier LLMs can now natively read/write neural network weights! 🧵
Blog: rikiyat.github.io/blog/weight-sp…

View quoted post on X

AI @DeepLearn007 Oct 29

Worth asking Google to comment and explain

Zhu Liang Verified account @paradite_ Oct 29

EXCLUSIVE BREAKING NEWS: google cheated in gemini 4 carbon benchmark reporting.

gemini 4 carbon didn't beat claude opus 5.5 in terminal-bench 4.0. carbon scores 65.2%, below opus 5.5 (66.4%).

but google let artificial analysis run tb4 benchmark, which put opus 5.5 below carbon by lowering opus 5.5 score.

at best this is sampling bias, at worse this is deliberate manipulation of public perception.

View quoted post on X