Yacouba Diarra @d1array

@oza_dev_ Formidable

The real pain of the AI revolution: my personal assistant saves me time, but replacing it every two weeks is a full-time job. It started when I saw the potential of having an AI agent as a personal assistant. I built something myself with OpenClaw. I was pretty happy until I realized I was spending too much time maintaining the thing that was supposed to save me time. So I switched to Claude. Great tool, but eventually it felt a little slow, a little expensive, and the experience of running it remotely wasn’t ideal. So I switched to Instinct. And I loved it. Suddenly, my assistant could book restaurants and manage my calendar. Exactly what I wanted. Except it needed a lot of permissions, and the AI security guy in me struggled to hand a young startup the keys to my entire life. So I switched to Muse. Connected it to everything. Taught it about me. Explained how I work, what I like, and what I need. Configured, corrected, refined. After a week or two, I was happy. Finally, an assistant that knows me, with everything working. And today, OpenAI released dot. Now I’m looking at my assistant the way I look at my iPhone the day after Apple announces a new one. I love technology. I love trying things early. But every switch means investing time all over again: rebuilding context, reviewing permissions, changing habits. Just as that investment starts paying off, something new arrives and the FOMO kicks in. What does being an early adopter even mean when, by the time you finish setting up the product, you’re already a legacy user? And when do we actually get to enjoy all the time we’ve saved? At this rate, I’m going back to a human assistant. At least she won’t become legacy by the time I finish introducing myself.

At some point in evolutionary history, matter learned to feel. We don't know how. It wouldn't be scientific to dismiss the possibility that it could happen again. (I don't believe it happens in current AI systems, because they don't have the charactistics I would expect to see in a sentient agent)

@WATeamSport Jamais été aussi hypé par un maillot!

…and confirm that everything works as expected on the flight by running it. In cases where you cannot run every possible scenario, like often in ML, you may occasional run into bugs that you can’t even explain, because you simply did not read the source code 🤣

AI has gotten good at coding, like really good. On the other end the pace of implementation is frightened now. You generate much more code than you can review, me personally I’ve been battling with that (today actually). I used to review every function written by AI assistants…

A habit from back when they could only implement simple module 3 years ago, now in big 2026 agents can implement a whole framework in 30 minutes, and I just can’t review 37+ code files before moving on with my experiments so I have to rely on unittest.

Inside you there are two wolves

Today is for the programmers 🍻

@DianboLiu That I understand. And you think paying the reviewers will encourage them to be more rigorous and making submitters pay will encourage them to develop more of those basic research skills? Or was that part just for trolling, specifically the Bitcoin part :)

@DianboLiu I imagine this overload makes it very tempting to use AI for review. Is that what you're worried about or the overload itself?

@DianboLiu Interesting take. Being new to publishing I spent some time thinking about how the current system even works. I submit my paper, and like a thousand other submitters I don't qualify for review and then a bunch of busy guys have to philanthropically review 20 papers each.

MIT published a brutally honest report on what AI is doing to students. A committee of professors and students spent five months studying how AI changed learning on campus, and the findings read like a warning to every university on the planet. Study groups are disappearing. Office hours are emptying out. Problem sets and take-home exams no longer prove anything, because AI can produce credible solutions to almost any written assignment in the undergraduate curriculum. Students who lean on chatbots lose mastery and confidence, and some slip into what the report calls cognitive surrender, reaching for AI at the first hint of struggle. The numbers are rough. 46 percent of surveyed MIT undergrads use LLMs daily. 90 percent worry about their own overreliance. Undergrads who feel AI makes them replaceable now outnumber those who feel it makes them capable. The committee's answer surprised me. They refused to fight AI with surveillance. The report calls AI detectors unreliable, says lockdown browsers feel like spying, and warns that policing students builds a classroom atmosphere of mutual distrust. Instead, MIT wants to rebuild education around the things AI can't replace. That means oral exams, semester portfolios, in-person project work, and a required social component in every subject. The report even floats the idea of rethinking grades entirely, since without a GPA to optimize, much of the incentive to cheat with AI evaporates. The committee warns professors against replacing undergrad research assistants with AI agents just because they're cheaper, because a university exists to grow people, not output. The most famous tech school on earth admitted the machines broke its way of teaching. Its answer is more humans, not more software.

Advancing cancer treatment: NUS scientists showed that brief, low intensity magnetic pulses can reprogram immune cells to eradicate breast cancer. This innovative approach avoids chemotherapy’s side effects and could potentially treat other cancers. news.nus.edu.sg/magnetic-pulse…

A very common sentiment I'm hearing among math students: "I don't want to become a mathematician anymore." I've seen this before, I think the generation of digital artists who came of age around 2022 will be the last. AI might not functionally replace mathematicians or artists, but it might well cause their near-extinction anyway, by crushing the spirits of the younger generations.

GPT 6 Astra is INSANE 🤯 It is a BEAST in 3D MODELING!! This is our first test, it blew our mind 🤯 It was able to rebuild a building infront of our house from just a few iphone photos in Blender in less than 30 minutes …

China is putting scientists on public screens and treating them like national celebrities, we are so fucked

Aw ni ce ! 🇲🇱 🇨🇮 🇧🇫 Le bambara entre dans l'ère de l'IA. Kooma — la première suite d'intelligence artificielle en bamanankan : voix, traduction, transcription, sous-titrage, doublage et clonage de voix. La voix du bambara, la puissance de l'IA. → kooma.ai #Bambara #Mali #BF #RCI #Bamako #Abidjan #IA