Developer, interested in many different tech fields. Working at Ubisoft Paris.
Joined February 2009
- Tweets18.3K
- Following216
- Followers211
- Likes24.1K
Manu_TechAndGames retweeted
I’m incredibly excited to share this video, in which I had the extraordinary privilege of mentioning the transformative impact of @OpenAI Codex and AI models on biomedicine, as well as what I’m working to develop using these technologies. I believe the video is beginning to air across various media outlets.
I hope it inspires others to explore the incredible possibilities AI is opening up for biomedical science and humanity! 🖖
I am truly grateful and thankful to OpenAI team for the opportunity and a tiny drop of contribution in this most important era of human existence.🙏
Some video games journalists don't have a clue what developing a game is.
Dear Tim Sweeney, you could learn a thing or two from the Xbox exec who posted so bad that he had to delete his account
aftermath.site/xbox-brandon-…
ALT tim
Ai for medicine is a real thing
Cancer patients with the same diagnosis often have very different tumor biology.
Molecular testing can reveal underlying differences, but that's slow and expensive.
Introducing Holon-S1, our platform to predict genes and biomarkers from routine pathology slides.
• beats two other models at predicting spatial gene activity in a controlled comparison
• has identified potentially novel biomarkers in breast cancer and colon cancer
• evaluated across 16 organ types and more than 15 scanners, using slides from both formalin-fixed, paraffin-embedded and fresh-frozen tissue.
We believe Holon-S1 has the potential to help researchers find new biomarkers in slides they already have, discover treatment targets, and bring more personalized cancer care within reach!
Manu_TechAndGames retweeted
Introducing Omnii-1, our most powerful genome language model, and a major step toward general biological intelligence
In a world first, Omnii-1 designed the first AI-generated bacteriophages to be administered to a human patient with an antibiotic-resistant infection under compassionate use
Manu_TechAndGames retweeted
I called this years ago.
I called it the intelligence optimum.
Once AI gets smart enough, more intelligence is longer a differentiating factor.
We're going to saturate intelligence, and then the competition will be on price, speed, reliability, extensibility, integration, and interoperability.
The hardware shops said this even longer ago: they expect intelligence to be fully commoditized by 2030.
Commoditized intelligence means that "SOTA model capability is table stakes"
OpenAI, Anthropic and other leading AI labs face a growing challenge: Better models aren't necessarily attracting more usage. A 10-point intelligence advantage now earns just 20% more developer traffic, down from nearly 69% in late 2025.
Full story: thein.fo/4hGJ9jY
Manu_TechAndGames retweeted
OpenAI, Anthropic and other leading AI labs face a growing challenge: Better models aren't necessarily attracting more usage. A 10-point intelligence advantage now earns just 20% more developer traffic, down from nearly 69% in late 2025.
Full story: thein.fo/4hGJ9jY
Manu_TechAndGames retweeted
A new frontier for on-device speech.
We're open-sourcing Phonon, a suite of TTS models under 50m parameters, setting a new standard for naturalness and accuracy across languages, with the broadest inference support to date.
Phonon-31m is available under Apache 2.0.
gradium.ai/phonon🧵
Manu_TechAndGames retweeted
Pocket TTS has been getting updates those last few days. While it was ~100M parameters, we release smaller versions, down to 1.9M: an 18.2M model with the quality of the original, a 6.9M model, and a 1.9M model that runs 103x faster than real time on one CPU core. All with voice cloning.
Most people really think they are rational and cartesian.
They even think it's a advantage.
Intuition is what makes us creative and efficient, rational comes after, to order what intuition said
It's only for US right?
Then it will make discrepancies between countries even bigger, leading to more conflicts, terrorism, immigration, and geopolitical issues.
Manu_TechAndGames retweeted
Sans même tenir compte des dégradations matérielles, ce weekend aura coûté au contribuable français environ 3 milliards *par an*, à perpétuité, en intérêts supplémentaires sur la dette (11bp de spread).
L'équivalent du salaire d'environ 60 000 profs de lycée supplémentaires.
Manu_TechAndGames retweeted
Marvell’s Dave Lazovsky is holding a tiny module that uses photonics to move high-bandwidth memory (HBM + DDR) as far as 50 meters away from the GPU.
this can sit up to 50 meters from a GPU instead of the usual 1.5 mm.
For the first time, memory no longer has to live right next to compute, letting AI clusters share massive capacity without being locked to the accelerator package.
Latency stays in the hundreds of nanoseconds, close enough for practical memory use.
Basically, Memory and compute now scale on separate axes instead of being locked together.
Manu_TechAndGames retweeted
为了方便使用AI做游戏时确定美术风格，我做了一个网站，收集了不少游戏的美术风格以及参考游戏，这个可以直接把风格名或风格参考发发给AI，让他知道要做什么。
gameartstyle.loopyhop.com
Manu_TechAndGames retweeted
Worth mentioning that at Epic, we’ve been experimenting and monitoring the development of AI production tools for several years. It feels like we’ve now hit an inflection point where they’re genuinely adding value for developers. So time to fully commit.
Interesting.
Sounds too good to be true
What if code could think while it runs?
Introducing Maxwell: the world's first Code Execution Model.
Maxwell lives inside programs as they run, and improves them! It's an entirely new class of model, sitting between code and hardware. No humans in the loop.
IMO Maxwell is the most direct way to achieve superoptimisation at scale.
Manu_TechAndGames retweeted
“We found DeepSeek V4.1 Flash 1017 beats GPT 6 Sol on our multi-step offensive cybersecurity task.”
READ THAT AGAIN.
the same cyber capabilities that OpenAI kept behind a gate got beaten by a Chinese open model you can run locally.
buckle up guys.
We found DeepSeek V4.1 Flash 1017 beats GPT 6 Sol on our multi-step offensive cybersecurity task. Sol was tested via the Daybreak trusted access program.
We have been building a cybersecurity benchmark based on simulated corporate networks. This is a valuable indicator for the cyber-risk a model poses if in the hands of a bad actor. We plan to make the full evaluation public in the coming weeks but wanted to share our first immediate finding here.
