ritz_labringue

6 ms·

1.
by ritz_labringue 1y ago
Well, it's really a VSCode extension that lets you run Codex CLI in the IDE. Not the "cloud" version of Codex... So GP is technically correct
2.
by ritz_labringue 1y ago
You’re right, I conflated two things. MoE improves compute efficiency per token (only a few experts run), but it doesn’t meaningfully reduce memory footprint. For fast inference you typically keep all experts in memory (or shard them), so V
3.
by ritz_labringue 1y ago
The short answer is "batch size". These days, LLMs are what we call "Mixture of Experts", meaning they only activate a small subset of their weights at a time. This makes them a lot more efficient to run at high batch si
4.
by ritz_labringue 1y ago
I don't think I agree with the premise. Sure there are lots of car accidents in absolute terms, but given how many people drive and how error-prone driving inherently is, most people are actually pretty decent drivers
5.
by ritz_labringue 1y ago
AI is really useful when you already know what code needs to be written. If you can explain it properly, the AI will write it faster than you can and you'll save time because it is quick to check that this is actually the code you want
6.
by ritz_labringue 1y ago
Harry Potter action figures trade almost entirely on J. K. Rowling’s expressive choices. Every unlicensed toy competes head‑to‑head with the licensed one and slices off a share of a finite pot of fandom spending. Copyright law treats that a
7.
by ritz_labringue 1y ago
I very much agree, and I think people who are in denial about the usefulness of these tools are in for a bad time. I've seen this firsthand multiple times: people who really don't want it to work will (unconsciously or not) sabota
8.
by ritz_labringue 2y ago
It does require writing good instructions for the LLM to properly use the tables, and it works best if you carefully pick the tables that your agent is allowed to use beforehand. We have many users that use it for every day work with real d
9.
by ritz_labringue 2y ago
It's not on par with o1, let alone o1-pro
10.
by ritz_labringue 2y ago
Why can't you just create a pool of databases and truncate all tables after each run ? That should be pretty fast
11.
by ritz_labringue 3y ago
researchers are paid 2x what engineers are paid at OAI, even if it's not the same job there's still one that is "higher level" than the other.
12.
by ritz_labringue 3y ago
> Create a blockchain-based AI-powered VR platform that allows pet owners to monitor and interact with their pets remotely, providing real-time feedback on their health, behavior, and emotions. Incorporate web3 technology to ensure secur
13.
by ritz_labringue 5y ago
Thanks ! Didn't know that. Isn't it also very expensive to run ?
14.
by ritz_labringue 5y ago
Doesn't it cost hundreds of thousands of dollars just to train GPT-3 ? If so, that seems like a good reason to use a "managed" GPT-3.
15.
by ritz_labringue 6y ago
I confirm that this product cannot be launched in France with this name
16.
by ritz_labringue 6y ago
Your parent is specifically talking about greenfield projects that are not customer facing. Alexa/Echo is customer facing.
17.
by ritz_labringue 6y ago
disclaimer: he was an investor in my startup He spends a lot of his money on giving back: 42, a free programming bootcamp ( https://www.42.fr/ ), Station F, a huge co-working space in Paris which hosts entrepreneurs for free
18.
by ritz_labringue 7y ago
Thanks. This makes sense.
19.
by ritz_labringue 7y ago
If you pay for the premium/metal plan, the support is stellar as well (or at least, it has been for me)
20.
by ritz_labringue 7y ago
Not sure why this is getting downvoted
21.
by ritz_labringue 7y ago
What does it have to do with the control key ?
22.
by ritz_labringue 7y ago
Not looking for a job, but the product looks awesome!
23.
by ritz_labringue 7y ago
Well, the website is actually pretty nice, maybe it's time for you to get a modern browser.
24.
by ritz_labringue 7y ago
This does not look like a reliable source of information at all.
25.
by ritz_labringue 7y ago
The keyboard is fixed on the 2018 MBP
26.
by ritz_labringue 8y ago
VS Code and Visual Studio are different beasts, though.
27.
by ritz_labringue 8y ago
I can't not mention this infamous SO answer: https://stackoverflow.com/questions/1732348/regex-match-open...
28.
by ritz_labringue 8y ago
I'm a junior developer in Paris (~1.3 years of experience). I make 54.5k/y "brut", which is basically a bit under 3000 after tax per month. 2500 seems a bit low to me, especially if you are in Paris and have 2+ years ex
29.
by ritz_labringue 8y ago
Unity itself (the engine) is written in C++. Game scripts are written in either C# or UnityScript iirc.
30.
by ritz_labringue 8y ago
Definetely not in France. FB Messenger is the primary way of messaging people here

More ›