bbor

68 ms·

871.
by bbor 2y ago
Yes, you should still trust cooperatives and syndicates. I am surprised they’re attempting such a brazenly disrespectful move, but in general, the people who started this company were self-avowed capitalists through-and-through; the fact th
872.
by bbor 2y ago
Any reasonable court would see right through “well we trained it for the public good, but only we can use it directly”. That’s not really a legal loophole as much as an arrogant ploy. IMO, IANAL
873.
by bbor 2y ago
Yes: non-profits usually have members , not shareholders. And, most importantly: non-profit charities (not the only kind of nonprofit, but presumably what OpenAI was) are legally obligated to operate “for the public good”. That’s why they’
874.
by bbor 2y ago
Totally agree that it’s “vestigial”, so it’s just like the nonprofits all the other companies run: it exists for PR, along with maybe a bit of alternative fundraising (aka pursuing grants for buying your own stuff and giving it to the needy
875.
by bbor 2y ago
Well, tbf, he’s one of the only ones left. There’s so much ethical/political/cultural attrition at these companies that even the company founded by attrition (Anthropic) looks like Theseus’ ghost ship these days. And Altman is, so
876.
by bbor 2y ago
We’re so interested in the public good that we can’t be hampered by the restrictive chains of legal obligations to the public! Don’t worry, we’ll just make rhetorical promises instead, it’s much more freeing. After all, who cares about
877.
by bbor 2y ago
They get to it if you scroll! I came to the comments with the exact same concern, but they eventually spell out the value-add in (IMO) very clear terms: First, we turned everything into data - a rich system of digital twins that enable
878.
by bbor 2y ago
Interesting question! I would guess that electrical signals to contract muscles occur IFF the muscles exert some force, regardless of whether the muscles are actually able to overcome other forces and physically contract. So I would guess t
879.
by bbor 2y ago
There's a link above with much more details: https://www.theverge.com/24253908/meta-orion-ar-glasses-demo... [The display] features Micro LED projectors inside the frame that beam graphics in front of your eye
880.
by bbor 2y ago
1. "The name Nazaré is the Portuguese version of Nazareth" ok y'all, I know you're working on edge tech, but lets cool the rhetoric down a little bit. Can't believe the PR-minded execs approved this choice. 2. &q
881.
by bbor 2y ago
Random side note: it appears that the scientific journal publishing the relevant paper is also publishing this science journalism piece, which makes me wonder why I’ve never seen that before. Why does Popular Science even exist anymore, r
882.
by bbor 2y ago
I love the moral direction, but this sadly doesn’t hold up to philosophical scrutiny. Is it murder to 1. Kill someone who’s about to kill someone? 2. Kill someone in a defensive war to defend your freedoms? 3. Kill someone by prioritizing t
883.
by bbor 2y ago
Yeah but this becomes very, very relevant for larger corpuses. It’s not just about pulling the right sized snippet in the end (really that could be done by an intermediate LLM call…), it’s also about supporting effective vector similarity s
884.
by bbor 2y ago
Wow, this is absolutely deviously simple, and drop-dead obvious in hindsight. Unless this person is copying someone else without knowing it, this little blog post is the most important LLM paper to come out in months! Assuming I understood
885.
by bbor 2y ago
Aww… definitely makes you feel for the man beneath the bluster. Most of it seems to be typical Marcus fare—basically that LLMs aren’t true AI yet are also dangerous yet are irreparably broken-but the segment on X-risk is the first time I’ve
886.
by bbor 2y ago
The results show that they may be a previously missing mechanism by which cells modulate their internal electrochemistry... "Our research shows that condensates influence cells well beyond direct physical contact, almost like they
887.
by bbor 2y ago
A) this is a buyer’s agent, not a seller’s agent. B) I don’t think you tell the agent your race, and traditional bias protections (namely doing the whole process without meeting in person) would still be in place AFAIU.
888.
by bbor 2y ago
I share your general concern, but tbf these people aren’t doing the obvious and basic “let a chatbot do it” thing, they’re more making a platform that involves a few specialized language models for making the expert-gated stuff accessible t
889.
by bbor 2y ago
Looks great, hopefully one day I’ll be able to use your services! Has all the features I’d be looking for. Out of curiosity, do you foresee regulatory trouble? Realtors are an extremely powerful lobby in the United States, and I wouldn’t be
890.
by bbor 2y ago
Well that's where it gets tricky -- what variables are you allowed to control for in this analysis? Not to spoil anyone, but this is what Soylent Green is about: people on the edges of society aren't actively coerced into euthen
891.
by bbor 2y ago
Thanks, “head of stock research”. Definitely a real job that’s not at all made up. I’m sure he did a lot of chin stroking and graph glancing to come to this technical decision! Someone should let all the scientists know that they’re wrong,
892.
by bbor 2y ago
Looks awesome! Not the kind of thing I need, but I see the clear value prop. And the site is slick af. Some random critiques: 1. You have a tiny typo in the “Uncover insights with AI” example (quire). 2. The docs are broken on mobile screen
893.
by bbor 2y ago
For the cynical among us, here’s a detailed paper of their upcoming proposals: https://sciencebasedtargets.org/resources/files/Aligning-cor... I was happy to see that they criticize carbon credits, but only when t
894.
by bbor 2y ago
Some of these are absolutely cursed suggestions for coding, namely Xanh, Syne, and Major. I love it! I would definitely throw the Monaspace family in there if they’re properly licensed, I’m something of an evangelist. Also there’s a free ve
895.
by bbor 2y ago
A) I don’t think evaluating individual clauses is an effective or clear rhetorical strategy for what they’re critiquing — all the “evaluations” are just “meh seems possible, sorta maybe” which doesn’t tell you much. B) lol, this is just a l
896.
by bbor 2y ago
Why. They already had absurd brand recognition, it wasn’t conflicting with anything other than maybe loosely Chrome, and it is rotationally symmetric. And to replace that with a typography logo that doesn’t even capture their common initial
897.
by bbor 2y ago
You’re (both!) getting into metaphysics without necessarily realizing it. He’s just saying that a machine that could learn any pattern—not a sub pattern of accidental actualities that it overfits on, but the real virtual pattern driving s
898.
by bbor 2y ago
You’re honestly convinced that he’s faking this level of futurism? I’m happy to see people call him wrong, maybe even defend him on a few point, but calling him dishonest on this central point seems irrational. Ditto for someone I persona
899.
by bbor 2y ago
Well put! If you/y’all haven’t heard, there’s a popular breakdown of “technical documentation” into four types, and this is one of the axes: https://nick.groenen.me/posts/the-4-types-of-technical-docum... People’s
900.
by bbor 2y ago
Wow, great catch. Something tells me he rolled this himself. Clearly he's trying to coin a term for personal legacy reasons, and I say godspeed. Holocene is a little vague, the information age is too entrenched to get anyone's

More ›