Armand du Plessis (@armanddp) 's Twitter Profile
Armand du Plessis

@armanddp

Obviously not a golfer. Mountain Trail Runner. Make the most of our one wild and precious life @pulseliveco 🕶 . Past: CTO - @hornet. [email protected]

ID: 3926311

linkhttps://pulse.tv calendar_today09-04-2007 17:33:07

14,14K Tweet

1,1K Followers

2,2K Following

Ryan Singer (@rjs) 's Twitter Profile Photo

I'm even more fond of Claude Code after this great convo with the creator and lead PM. Refreshingly practical and down-to-earth. No AGI nonsense. (I've used Claude Code more than other tools for coding. Especially to learn how the existing system works.) youtu.be/zDmW5hJPsvQ?si…

Andrej Karpathy (@karpathy) 's Twitter Profile Photo

An attempt to explain (current) ChatGPT versions. I still run into many, many people who don't know that: - o3 is the obvious best thing for important/hard things. It is a reasoning model that is much stronger than 4o and if you are using ChatGPT professionally and not using o3

An attempt to explain (current) ChatGPT versions.

I still run into many, many people who don't know that:
- o3 is the obvious best thing for important/hard things. It is a reasoning model that is much stronger than 4o and if you are using ChatGPT professionally and not using o3
delian (@zebulgar) 's Twitter Profile Photo

Kinda hard to operate the station without a vehicle to bring crew and cargo to it And kinda hard to de-orbit the station safely without using Dragon to bring it down Elon has Starlink to fund the company and wants to focus on Starship/Mars, so why would he back down?

Armand du Plessis (@armanddp) 's Twitter Profile Photo

So backwards from SANRAL to not accept tap payments and insist on physical cards. They must be inconveniencing and frustrating thousands of motorists daily. Who still carry cards even?

Andriy Burkov (@burkov) 's Twitter Profile Photo

Apple did more for AI than anyone else: they proved through peer-reviewed publications that LLMs are just neural networks and, as such, have all the limitations of other neural networks trained in a supervised way, which I and a few other voices tried to convey, but the noise

Armand du Plessis (@armanddp) 's Twitter Profile Photo

“The coming days will test whether Iran’s 85 million people can seize this moment to build something new… or whether the aftermath becomes a new tragedy.”

Andrej Karpathy (@karpathy) 's Twitter Profile Photo

Nice - my AI startup school talk is now up! Chapters: 0:00 Imo fair to say that software is changing quite fundamentally again. LLMs are a new kind of computer, and you program them *in English*. Hence I think they are well deserving of a major version upgrade in terms of

Mike Taylor (@hammer_mt) 's Twitter Profile Photo

The bitter lesson: hard coding and hand holding AI systems always succeeds in the short run but fails in the long run compared to just throwing more data and compute at the problem. Tesla just achieved the same level as Google but with immeasurably more cars on the road, far less

DHH (@dhh) 's Twitter Profile Photo

Sevens years later, and I must admit that Elon's vision, tenacity, and sheer willpower actually delivered. I should not had focused so myopically on the present in 2018, but on where it might land in 2025. Progress demands faith in the future. Mine has been reset!

Steve (Builder.io) (@steve8708) 's Twitter Profile Photo

ive stopped paying attention to LLM benchmarks the reason why claude code is blowing up is because of how effective anthropics models are at using tools the other models, including those that win all the benchmarks, don’t even come close

ive stopped paying attention to LLM benchmarks

the reason why claude code is blowing up is because of how effective anthropics models are at using tools

the other models, including those that win all the benchmarks, don’t even come close