bbor

33 ms·

421.
by bbor 2y ago
Execution is incredible, and the idea is far from bad, but I must I say that it is certainly hilarious . Giving serious “we’ve figured out a way to more durably connect microservices!” energy. Ngl, I see a non-zero chance of this kind of a
422.
by bbor 2y ago
This is a lively debate in linguistics, so while it’s not objectively true (i.e. it depends on how you want to use the word), it’s much more justified than you’re implying here. As others have said, it’s about distinguishing the kind of c
423.
by bbor 2y ago
I definitely think you’re touching on some exciting possibilities, but adding a language model at this early stage would endanger the goal of this particular research: proving that the compositionality exists in the first place. If there wa
424.
by bbor 2y ago
Really great, succinct way to make this point. Here's an NGRAM of mentions of these brands in the English Fiction corpus, 1860-2025 -- Ferrari dominates until ~1970, when Porsche gains dominance. Obviously, Koenigsegg is barely on the
425.
by bbor 2y ago
I had fun writing the reply too. I enjoyed the piece, regardless! Certainly don't take my disagreement as a reason not to do more shenanigans :)
426.
by bbor 2y ago
lol, obviously, but there's also definitely part of me that thinks this would've worked better as a bsky post than a whole serious-sounding blog analysis. The core complaint is made explicit in the postscript: ...let's no
427.
by bbor 2y ago
I really don’t know what to do with posts like these — a startup announced today supposedly beats all other OSS models out of the gate?? I mean, they have screenshots of tables and stuff so they’re presumably not outright lying, but I’ve
428.
by bbor 2y ago
Product-wise this seems to fall into the "sounds great, if it a) works well and b) is cheap enough to use regularly" category, so instead I'll leave some more general musings on the company for my fellow nosy outsiders: After
429.
by bbor 2y ago
Evolution is a stochastic process operating almost entirely in the shadowy past, so the scientifically-responsible answer to this “why?” question is “we don’t know for sure”, I think we can all agree. Moving past that to speculate though,
430.
by bbor 2y ago
I'm guessing you know this, but for those new to the terms: what this post describes as the "VC route" and the "bootstrap route" are usually referred to as "growth businesses" and "lifestyle businesse
431.
by bbor 2y ago
Plus, at a certain point we should probably ask what we're even doing here. At the same time and using the same ostensible "pro-environment" framework, we: 1. refuse to engage in biome modification to save soon-to-be-homeless
432.
by bbor 2y ago
It's not filling in the blanks that's impressive, it's meaningfully combining them all into an objectively unique narrative, building upon those blanks at length. Definitions are always up for debate on instrumental grounds,
433.
by bbor 2y ago
Disregarding the (common!) assumption that AGI will consist of one monolithic LLM instead of dozens of specialized ones, I think your comment fails to invoke an accurate, consistent picture of creativity/"truly new" cognition
434.
by bbor 2y ago
IMHO there's something to be said for innovative, paradigm-defining[1] thinkers being more likely to accept frameworks that we in hindsight recognize as definitively disproven. Not to say alchemy was exactly an open question in Newtoni
435.
by bbor 2y ago
I’m glad we have this result to confirm what’s obvious to some of us and completely absurd to others, but it’s also worth pointing out that the Turing test was never meant to be a literal test. He invoked the “Imitation Game” to make a phil
436.
by bbor 2y ago
I guess I'm saying that we also can't observe any proof for the theory that anything exists beyond the solar system instead of a giant screen run by aliens built to trick us. I appreciate the empirical mindset, but taking it this
437.
by bbor 2y ago
Not to disagree with your justified Socratic questioning, but this sparked my interest and I figured I'd share my (novice!) TIL: it appears the IAU uses a coordinate system based on our solar system on January 1st 2000, 00:00 AM, or &q
438.
by bbor 2y ago
I'm not sure I understand this response -- basically all of cosmology involves theories based only on somewhat subjective interpretations of sensor data targeting far-off objects. So I don't necessarily disagree . But what could
439.
by bbor 2y ago
Interesting, thanks for elaborating! Still not sure I understand sadly, but maybe it’s just different worldviews on how to play language games. Like, you say we don’t need to teach nuance, and that animal is a useful term — but then what do
440.
by bbor 2y ago
I saw this on bsky a few days ago and was fascinated, so glad to see it made its way here! I had this question that went unanswered, so hopefully there's some experts/fellow musers here: Why grow up if you don't need photosy
441.
by bbor 2y ago
I'm a huge cladistics fan, but insisting that "animal" isn't a useful term is pretty out there, IMHO. Certainly we can (and do!) teach cladistics in high school biology, but does your ideal world have baby books titled w
442.
by bbor 2y ago
Look for the ways that AI works, and it can be a powerful tool. Try and figure out where it still fails, and you will see nothing but hype and hot air. Perfectly put, IMO. I know arguments from authority aren't primary, but I
443.
by bbor 2y ago
I don’t disagree with any of your particular points, but I think you’re missing the forest here: their argument is primarily based in empirical results, not a theoretical framework/logical deduction. In other words, they’re trying to
444.
by bbor 2y ago
Yup -- asking a chatbot to create a novel in one shot is very similar to asking a human to improvise a novel in one shot.
445.
by bbor 2y ago
Whelp. That's terrifying.
446.
by bbor 2y ago
Very, very cool -- thanks for posting! Quanta is yet again proving itself to be the very best casual science publication in the English-speaking world. This is a short one, but as always, clean, original graphics + big pictures of the parti
447.
by bbor 2y ago
Heh, gotta love HN -- we're all just musing, and in comes a founder of the biggest startup in this space to share their personal experience! Thanks for commenting, interesting stuff. I think you hit the nail on the head with "Link
448.
by bbor 2y ago
I have to disagree, as HN's resident anti-Display-Ads evangelist: the problem is the display ads! Display ads mean influencers, and LinkedIn messed up specifically by paying influencers extra if they got people to click "See More&
449.
by bbor 2y ago
It is the law in California, and it is certainly possible at scale, just not cheap or trivial . Presumably, given the intense focus of the state government on their company in particular, they have long since implemented a compliant dele
450.
by bbor 2y ago
I'd say this is better for those closer to the engineering side of the science<->engineering continuum, for sure. For AI researchers proper, I happen to strongly believe in symbolic approaches, so I'd say it's more of a

More ›