frenchmajesty

7 ms·

1.
Inference, Diffusion, World Models, and More – YC Paper Club [video](youtube.com)
3 points by frenchmajesty 3mo ago|0 comments
2.
How to Scale Your Model – A Systems View of LLMs on TPUs (2025)(jax-ml.github.io)
2 points by frenchmajesty 3mo ago|0 comments
3.
Show HN: The CTO Game – Scale your infra in real-time, under pressure(thectogame.com)
3 points by frenchmajesty 7mo ago|0 comments
4.
by frenchmajesty 9mo ago
Cool concept! However, the fact that it's on a timer and you can only try the next even in 1m is a killer feature (in a not-good way). Same for not being able to view the leaderboard
5.
by frenchmajesty 9mo ago
Historically this has never flown. There's only 1 billionaire for every what, 10M people, 100M ? They are outnumbered. At some point the underclass may band together and kick off a revolution to overthrow the ruling class. Kings have g
6.
All of My Employees Are AI Agents, and So Are My Executives(wired.com)
16 points by frenchmajesty 10mo ago|11 comments
7.
Scheduling in Go (2018)(ardanlabs.com)
1 points by frenchmajesty 10mo ago|0 comments
8.
by frenchmajesty 10mo ago
Yea it ends on that date. I cashed out with $622K
9.
by frenchmajesty 10mo ago
I do get answers. The ones who typically said can't disclose are you full of shit. Crack teams know they're cracked and they will tell you. "I'm a MIT-trained security engineer; I'm an ex-VP at Paypal, etc..."
10.
by frenchmajesty 10mo ago
Damn. Now I understand VC's a little better. Some deals I saw "damn the traction is off the chart" and I didn't even care to ask any question I just threw money at them. I saw one startup named Salesforce, a CRM. I gues
11.
by frenchmajesty 10mo ago
Fun game! My investing strategy is to flush out quickly who's the team, what's the traction, what's unique insight into the market. Bullshitters have trash or no answers. Killer companies have one of those 3 that is usually v
12.
by frenchmajesty 10mo ago
OP here. I should've added that clarification but it was to say as someone with the "most to gain" financially or professionally from AI becoming more important, I'm still against this content trend and find it is unprod
13.
by frenchmajesty 10mo ago
The ideas submitted still go through you where you approve/reject.
14.
by frenchmajesty 10mo ago
Good platform. A similar one was made I used to use back in ~2020 but eventually I think it died. The best addition would be to have a simple form where people can submit ideas for questions. It'll keep your site from being stale and f
15.
I am just sooo sick of AI prediction content, let's kill it already(verdikapuku.com)
74 points by frenchmajesty 10mo ago|71 comments
16.
You need to become a Full Stack person; Product Engineer roles on the rise(den.dev)
2 points by frenchmajesty 10mo ago|0 comments
17.
Polish to be the most effective language for prompting AI, new study reveals(euronews.com)
4 points by frenchmajesty 10mo ago|1 comments
18.
The Learning Loop and LLMs(martinfowler.com)
1 points by frenchmajesty 11mo ago|0 comments
19.
RSF forces in Sudan's civil war are preparing for mass genocide(economist.com)
39 points by frenchmajesty 11mo ago|19 comments
20.
Ask HN: Mamdani is poised to become new mayor of New York. How do locals feel?
5 points by frenchmajesty 11mo ago|8 comments
21.
by frenchmajesty 11mo ago
You are looking at it wrong. Meta is a business. You know what they sell? Ads. In fact, they are the #1 or #2 place in the world to sell an ad depending on who you ask. If the future turns out to be LLM-driven, all that ad-money is going to
22.
Twitter/X announces to start auctioning rare user handles for millions(engadget.com)
7 points by frenchmajesty 11mo ago|3 comments
23.
by frenchmajesty 11mo ago
I get a Heroku error trying to view the site.
24.
by frenchmajesty 11mo ago
Hey OP here. You're not wrong! Leaving aside the philosophical debate (isn't all form of capitalist participation selfishly motivated?), the main motivator was to help me and my friends with a problem we struggled with. Many solo
25.
by frenchmajesty 11mo ago
Hey OP here. The use-case is to give an Agent the ability to post on my behalf. It can use these class labels to figure out "what are my common niches" and then come up with keyword search terms to find what's happening in th
26.
by frenchmajesty 11mo ago
OP here. I agree! I should've called out why I did _not_ follow that approach as many others have commented the same. The main reason why is that I needed the classification to be ongoing. My system pulled over thousands of tweets per
27.
by frenchmajesty 11mo ago
OP here. We embed both the label AND the tweet. So if tweet A is "I love burgers" and tweet B is "I love cheeseburgers", we ask in our vector DB if we have seen a tweet before that is very similar to B? If yes, we skip L
28.
by frenchmajesty 11mo ago
OP here. I agree with you. For production use we use VoyageAI which is usually 2x faster than OpenAI at similar quality levels (p90 is < 200ms) but we're looking at spinning up a local embedding in our cloud environment, that would
29.
by frenchmajesty 11mo ago
Op here. Yes that's right. We do also insert the current text embedding on misses to expand the boundaries of the cluster. For instance: I love McDonalds (1). I love burgers. (0.99) I love cheeseburgers with ketchup (?). This is a bad
30.
by frenchmajesty 11mo ago
OP here. Yes that works too and get you to the same result. Remove risks for bias but the trade-off is higher marginal cost and latency. The idea is also that this would be a classification system used in production whereby you classify dat

More ›