Yacouba Diarra

And what’s beyond the infinity ?

[i🅱️RIZZLE 🥶 @realcleanbrr]

Computer science lol

[Yacouba Diarra @d1array]

@spisallyouneed Oh so, electricity is affordable and ubiquitous. Talk, about news huh

[Yacouba Diarra @d1array]

@VictorTaelin Yeah, I think that's just the defaullt mode of any sane individual

[Geoffrey Hinton @geoffreyhinton]

Penguin has moved up the publication date of my new book with Patchen Barss called "Smarter Than Us". The book explains how AI works for a non-technical reader and then talks about the myriad risks and what to do about them.

[Yoshua Bengio @Yoshua_Bengio]

Mila’s annual supervision request process is open for MSc and PhD applications for Fall 2027 admission. While I won’t be taking on new students for supervision this cycle, I encourage you to apply to join Mila’s brilliant deep learning research community.

For more information and guidelines, you can visit the link below or contact the @Mila_Quebec team directly.

https://t.co/vilZxgxbMg

[Yacouba Diarra @d1array]

@oktra_dev Prendre volant de quoi? Avec ton argent grand frere ou bien?

[Tesla @Tesla]

Real life, baby

[Yacouba Diarra @d1array]

😂

[Taelin @VictorTaelin]

I'd sell my kidney for a smart glasses that put live subtitles on people's English speech :')

[DJELIA @djeliia]

Pas développeur?

Sur https://t.co/1JPB4zW7T2, enregistrez quelques secondes de bambara : le texte apparaît. Collez une phrase en français : Moussa la lit en bambara.

Sans une ligne de code, dans votre navigateur 👇
https://t.co/1JPB4zW7T2

👨‍💻Developpeurs?

Votre code tourne déjà avec le SDK #OpenAI ?
Alors vous êtes à 3 lignes du #Bambara #Dioula.

1. api_key → votre clé @djeliia
2. Docs → https://t.co/oBeCned562
3. model → sunjata-1, jifili-1 ou banjugu-1

Clé #API en quelques secondes.

#DJELIA #IA #Africa

[NARUTO・BORUTO【原作公式】 @NARUTO_kousiki]

岸本斉史が描く
『NARUTO-ナルト-』新作漫画
⠀
『NARUTO-ナルト- SHADOW RISING』

続報は #ジャンプフェスタ 2027
『NARUTO-ナルト-』 スーパーステージにて𖣘
⠀
#PROJECT_RASENGAN #NARUTO

[Massachusetts Institute of Technology (MIT) @MIT]

We’ve lost an icon. Margaret Hamilton, a computer science pioneer best known for her work at MIT leading development of the flight software for the Apollo missions to the moon, has died at 90. Read about her legacy via MIT News: https://t.co/ODkbdIsoHw

[Yacouba Diarra @d1array]

@MadridSports_ Funny how Courtois and Konate are also enjoying themselves but nobody seems to care

[Yacouba Diarra @d1array]

@oza_dev_ Formidable

[Itamar Golan 🤓 @ItakGol]

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.

[François Chollet @fchollet]

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)

[Yacouba Diarra @d1array]

@WATeamSport Jamais été aussi hypé par un maillot!

[Yacouba Diarra @d1array]

…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 🤣

[Yacouba Diarra @d1array]

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…

[Yacouba Diarra @d1array]

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.