blueblimp

7 ms·

1.
by blueblimp 7mo ago
Seems bizarre. It's not like companies didn't want to sell it--they'd prefer to have the revenue. This is just kicking them then while they're down. I wonder if it will reduce risk-taking since it increases the downside
2.
by blueblimp 1y ago
The inter-city travel was my favorite part of EverQuest. (The rest of the game, I didn't find too interesting.) The level of challenge was about right: if you looked at maps and planned your route, you could generally get to where you
3.
by blueblimp 1y ago
I enjoyed browsing through it. One comment: the "philosophy and art" row is missing anything after 1890.
4.
by blueblimp 1y ago
> Today, 1,165 Bob Ross originals — a trove worth millions of dollars — sit in cardboard boxes inside the company’s nondescript office building in Herndon, Virginia. This seems like a bit of a waste given that there's demand for the
5.
by blueblimp 1y ago
I feel it too: - Plenty of em-dashes - "you're absolutely right" - "They're X, not just Y"
6.
by blueblimp 2y ago
And the most important thing about PCC in my opinion is not the technical aspect (though that's nice) but that Apple views user privacy as something good to be maximized, differing from the view championed by OpenAI and Anthropic (and
7.
by blueblimp 2y ago
> Yet this approach is obviously much better than what’s being done at companies like OpenAI, where the data is processed by servers that employees (presumably) can log into and access. No need for presumption here: OpenAI is quite trans
8.
by blueblimp 2y ago
This sort of convenient semi-arbitrary extension of a partial function is ubiquitous in Lean 4 mathlib, the most active mathematics formalization project today. It turns out that the most convenient way to do informal math and formal math d
9.
by blueblimp 2y ago
> many of my math friends consistently slept 9–10 hours a day. Anecdotally, I've noticed an association between long sleeping and math ability in particular, so this doesn't surprise me. I wonder if it's been studied scien
10.
by blueblimp 2y ago
In Meta's case, the problem is that they had been given the go-ahead by the EU to train on certain data, and then after starting training, the EU changed its mind and told them to stop.
11.
by blueblimp 2y ago
Video of Hyperia: https://www.youtube.com/watch?v=0wUKY3HBw2U A fun point in the article: Burton got into the industry via Rollercoaster Tycoon experience: > Curiosity became an obsession in his teens, when he started to
12.
by blueblimp 2y ago
I agree: a well-designed educational course can feel like a video game in some ways, in that you're learning at a high, consistent rate.
13.
by blueblimp 2y ago
My favorite blog post on tips for finishing is this classic by Derek Yu (of Spelunky fame): https://makegames.tumblr.com/post/1136623767/finishing-a-gam... .
14.
by blueblimp 2y ago
That was strange. I had to reread it to confirm that he's just talking about some random kid looking at the publicly-available game data files, not an employee or anyone else with special access.
15.
by blueblimp 2y ago
All sorts of technology can be used secretly to assist criminal enterprises. Cars, computers, pencils, electricity, etc. It's unfair to hold LLMs to a higher standard than what applies to nearly everything else.
16.
by blueblimp 2y ago
Yep. Advanced military technology tends to favor countries with high GDP per capita, which are mostly liberal democracies.
17.
by blueblimp 2y ago
The license allows to reproduce/distribute/copy, so I'm a little surprised there's an approval process at all.
18.
by blueblimp 2y ago
Qwen1.5-72B-Chat is dominant in the Chatbot Arena leaderboard, though. (Miqu isn't on there due to being bootleg, but Qwen outranks Mistral Medium.)
19.
by blueblimp 2y ago
The Bradley-Terry model is essentially the same thing as Elo. https://lmsys.org/blog/2023-12-07-leaderboard/ > This model actually is the maximum likelihood (MLE) estimate of the underlying Elo model assuming a
20.
by blueblimp 3y ago
This was definitely easier to follow. Since they're building a special-purpose accelerator for a certain class of models, what I'd like to see is some evidence that those models can achieve competitive performance (once the hardwa
21.
by blueblimp 3y ago
> When you say Claude is worse than Mistral Medium are you going by the Chatbot Arena Leaderboard or some other benchmark? I'm mainly going by the arena leaderboard, but it's also true in my limited experience with the two. (I
22.
by blueblimp 3y ago
The tech is legitimately impressive and exciting, but I couldn't help but chuckle at the revenge of the Scunthorpe problem: > It looks like the safety filter may have taken offense to the word “Cocktail”!
23.
by blueblimp 3y ago
For getting a flavor of training LLMs without needing to be at one of the pre-training companies, it's very accessible to fine-tune a relatively small open source LLM such as Mistral 7B. (There are many tutorials.)
24.
by blueblimp 3y ago
I wonder what's happening to all that money. Back when they originally released Claude, they were second only to OpenAI as far as chatbot models were concerned. Although Claude wasn't as smart as GPT-4, it had a more pleasant writ
25.
by blueblimp 3y ago
It's still an exceptionally poor privacy policy compared to pre-LLM online services. Compare with the Google Docs privacy policy, for example: https://support.google.com/docs/answer/10381817?hl=en > Google
26.
by blueblimp 3y ago
Yep. Unfortunately, all the major LLM APIs do surveillance on your prompts and responses.
27.
by blueblimp 3y ago
Yeah, I'm curious what's going on. Canada seems to be the only developed country without Bard at this point. (US, UK, EU, Australia, New Zealand, Japan, South Korea, Taiwan... all there.)
28.
by blueblimp 3y ago
I don't know what Twitch's goals are here, but if they want to get rid of the right-up-against-the-line content, then maybe clear, consistently-enforced rules are actually detrimental, because it just invites rules lawyering like
29.
by blueblimp 3y ago
I question the article's framing of CPUs as "universal" and GPUs as "specialized". In theory, they can both do any computation, so they differ only in their performance characteristics, and the deep learning revolut
30.
by blueblimp 3y ago
> I regard a physics education as a modern-day ‘liberal arts’ technical degree. Some anecdotal evidence for this: https://www.dwarkeshpatel.com/p/dario-amodei#details > Dario Amodei: We have generally found that

More ›