HN user
building - @fanahova / latent.space
Congrats!! I remember growing up in Italy and really really wanting to attend Hacker School but couldn’t afford to. It resonated a lot with me and make me want to keep learning; I am sure you are still having that impact on others today, it’s great to see it’s still around!
NVIDIA already designs most of their chips using AI. Why would you assume it's meaningless marketing?
How many more years do you want to wait? I own a store and we have not really seen any dip in attendance. There are definitely lots of older players churning, but lots of new ones being brought in as well. It's clear the game is being Commander-ified and lots of folks who were more interested in constructed / competitive Magic aren't happy, but it doesn't mean the game will die or anything like that.
Final Fantasy, LOTR, Avatar, are some of the best selling sets in the history of Magic. As much as I don't like some of the spiderslop, there is no real argument against printing more of it. The Hobbit set is going to be a complete blowout success.
There is plenty of precedent being written here. It does not seem to be the case at all for the average use of this technology.
Solving merge conflicts on text-based files is infinitely easier than binary-based. It's not a useful comparison.
Are you saying there are only 30 million people employed in white collar jobs in the world?
Everything in this comment is wrong lol
You know it's a EU study because they bring up "AI patents" in the first 2 minutes of it, as if they mean anything
People can write horrible PRs manually just as well as they do with AI (see Hacktoberfest drama, etc).
"LLM Code Contributions to Official Projects" would read exactly the same if it just said "Code Contributions to Official Projects": Write concise PRs, test your code, explain your changes and handle review feedback. None of this is different whether the code is written manually or with an LLM. Just looks like a long virtue signaling post.
I agree, that's why I was trying to point out that saying "if a person did that we'd have a word for them" is useless. They are not people, and people don't behave like that anyway. It adds nothing to the discussion.
Are you saying that every piece of code you have ever written contains a full source list of every piece of code you previously read to learn specific languages, patterns, etc?
Or are you saying that every piece of code you ever wrote was 100% original and not adapted from any previous codebase you ever worked in or any book / reference you ever read?
One co trying: https://www.system.com
Because they make $60B/yr on advertising and car sales is a very valuable ad market.
not truly groundbreaking foundation models.
Where is any proof that Yann LeCun is able to deliver that? He's had way more resources than any other lab during his tenure, and yet has nothing substantial to show for it.
The structure of each section gives away that it's mostly AI even without having to read the actual words. I'm sure it was AI + writer, but there's something about ending each section with 3-4 short, question-like sentences that is strongly AI. This is the same format as the successful LinkedIn slop so maybe it's not AI and just algo-induced writing.
The equivalent of "If you have to ask, you can't afford it" here is "If you have to ask, you shouldn't do it".
Imagine being able to retire at 40 and do whatever you want. If you weren't stupid, your health should be good enough.
Do you really believe people who have health issues at an early age are simply stupid?
You really think the amount of savings in 401ks is the same size as the GDP of the whole country?
You can pay more. It's unlimited (sorta) through API at API pricing.
The non-tinfoil hat approach is to simply Google "Boston demographics", and think of how training data distribution impacts model performance.
The data set used to train CheXzero included more men, more people between 40 and 80 years old, and more white patients, which Yang says underscores the need for larger, more diverse data sets.
I'm not a doctor so I cannot tell you how xrays differ across genders / ethnicities, but these models aren't magic (especially computer vision ones, which are usually much smaller). If there are meaningful differences and they don't see those specific cases in training data, they will always fail to recognize them at inference.
The problem is that: - These are not really super computing cluster in LLM terms. Leonardo is a 250 PFlops cluster. That is really not much at all. - If people in charge of this project actually believe R1 costs $5.5M to build from scratch, it's already over.
The Mtg Pro Tour is free of entry and has a $500,000 prize pool. Tournaments encourage people to buy cards.
Where do you think prize support for tournaments would come from if no one had to buy the cards?
That is not true. Try playing a $0.30 Underground Sea at Eternal Weekend and see how many rounds it takes before you get caught. Old cards have specific hues, imperfections, etc, that are not replicable in modern proxies. I have some Legacy proxies for local events that are proxy-friendly, and literally the first game I played someone noticed as soon as I put the card down that it was fake because it was printed way too well.
Avg player doesn't buy a few thousand cards at a time. If you buy a high value card from a random seller you should always check it unless you trust them from references.
What is missing in the context here is that the cards mentioned in this article are not actually real. They never existed, and therefore they are not "counterfeits" of a real one, they are just made up. Someone just claimed to know someone that had playtest cards from back in the day. They are not a commercial product.
See here for a bit more background: https://www.cgccards.com/news/article/13347/
We covered that too: https://www.latent.space/p/ai-engineer
If you joined 8 months ago it might be hard to recognize. I've been on HN for more than a decade and the quality of discourse has drastically lowered in quality especially in the last 3-4 years. This is a problem with the broader web, not just HN. Tech / startups is now a mainstream topic that attracts a lot of people who are not really in the weeds and are just able to write surface level comments.
Regarding the name, open is just a word. Apple doesn't sell apples. The company never promised to open source every model, only to make them accessible to the public, so you're arguing semantics that lead to no improvement in the technical conversation.