122 points by ray__ 3 days ago | 78 comments | View on ycombinator
abahlo 3 days ago |
epgui 3 days ago |
mjdiloreto 2 days ago |
Anyone familiar with Dr. John Vervaeke's work will find another notion, relevance realization (RR), fits nicely.
The thing I like about RR over "taste" when talking about AI is that it captures what the taste metaphor does not - that there is a missing _relevant_ dimension of consideration from the LLMs output.
"Hey you got the code working but this 4000 line PR will never be approved." "You spent tokens hardening the API, when this entire project is just a local test that will never hit production". Etc, etc, ad nauseum.
Vervaeke reveals why in his lecture series "Awakening from the Meaning Crisis".
Fundamentally, RR cannot be achieved through reason alone. Full stop. Vervaeke uses this fact to point out the irrationality of the "rationalist" movement in order to highlight its dangers (which many would agree the effective altruism crowd has thoroughly proven him right about), but the same framing and argument can be used to explain why your AI has bad taste.
There is no way to enumerate the set of dimensions that are or are not relevant in a given real world context. The detail of the universe, combined with the set of actions an agent does (and does NOT!) take leads to a combinatorial explosion of complexity.
AI agents need to be told what to do, and even then, they will only stochastically sample what you tell them.
Anthropic tells Claude lots of things about Rust, for example, and things to look out for in security context, but they fundamentally cannot put enough information in to grant the agent the ability to wisely choose what is and is not relevant. You can only ever consider _some_ relevant things.
I don't personally think it is impossible for computers to perform RR, it's just that today's agents do not, which is why its taste and yours differ.
aDyslecticCrow 3 days ago |
A video would have been been simpler and easier to navigate. A video would have been faster as well; even when scrolling ahead i would have liked a 150% option.
decidu0us9034 3 days ago |
Luke2642 3 days ago |
tsunitsuni 2 days ago |
achairapart 3 days ago |
spieke 2 days ago |
Turns out, if you have Animations disabled in your Windows settings (aka prefers-reduce-motion), you see a simple text web page.
xydac 3 days ago |
lhk931122 2 days ago |
jrdnfrd 3 days ago |
Maybe right now that's only true in a few domains like math, programming, language translation, maybe some writing task? But soon it will be true in many more domains.
Given that we can't hope to out-execute the LLMs on the things they are elite at, what is left for us? This site wants to reserve "taste" for the humans, but I'm not sure LLMs won't learn "taste" too.
The best I've got right now is that humans live and interact with the real world in ways that disembodied LLMs can't (yet). Maybe they will be the theoreticians and we the experimentalists?
tiborsaas 3 days ago |
FrameworkFred 3 days ago |
irickt 1 day ago |
Powerful tools reward the people who showed up sharp, iterate, spend extra time on the vision. They expose the people who didn't. AI is not the great equalizer, it is the great amplifier.
meshweaver 3 days ago |
aalazz 2 days ago |
xnx 3 days ago |
a3w 3 days ago |
fg137 3 days ago |
0x4e 3 days ago |
needfish 2 days ago |
gregrobson 3 days ago |
designer_bri 2 days ago |
etherlink 2 days ago |
kqp 2 days ago |
It’s “in plain text” and also the most whimsical excess you’ve seen today. Both of these things are “very human”. No further thought, your honor.
It’s a manifesto and notes-to-self, not a thinkpiece. No further thought, your honor.
It’s not a thread or a deck. You’re reading this on a website. No further thought, your honor.
“Symmetry is a rule. Break it. Three-act structure is a rule. Break it. "Don't shoot into the sun" is a rule. Break it.” This is itself a bizarre symmetrical trifecta. No further thought, your honor.
Near the beginning it’s “last edited at the kitchen table, coffee #2”, near the end “last touched, kitchen table, late”. No further thought, your honor.
I don’t care to sacrifice more of my life to the rest of this, I’m sure there’s worse to dig for.
Propaganda works not by stating the message, which makes it only something a person said, but by assuming it, which makes it something you’ve figured out about the world. Of course, this is very shallow, and accidentally says the weaponized assumption plainly, too: “the machines are very good and getting better every day”.
If they have to resort to this, and if this is the best they can do at it, I’m really not worried about “figuring out what’s left for humans”.
hamuraijack 3 days ago |
vunderba 3 days ago |
Seriously - these are so bad I almost want to believe this was done using Stable Diffusion 1.5 on purpose. Or maybe they used a Simpson LoRA.
dmitrygr 2 days ago |
smusamashah 3 days ago |
cykros about 22 hours ago |
Really wishing we could just get back to creating sites that work in w3m. I get we want graphics, but it loads images just fine in a framebuffer.
I literally can't be bothered to figure out what information they were trying to convey with me because this page was simply unusable.
dsjoerg 3 days ago |
happyopossum 3 days ago |
SwellJoe 3 days ago |
It's also a high bullshit design...a lot of random shit happening for no reason other than to make it hard to read and weird to navigate.
Maybe good intentions, but I can't get past the surface level stuff that's very off-putting.
makerofthings 2 days ago |
what 3 days ago |
withinrafael 2 days ago |
tomaskafka 2 days ago |
MetaMalone 3 days ago |
mkgiga 2 days ago |
mahboi 2 days ago |
gcanyon 3 days ago |
That page is anything but plain text...
workfromspace 2 days ago |
queenkjuul 2 days ago |
It's very pretty but after 30 seconds i gave up so i have no idea what the point was
dreamcompiler 2 days ago |
(Too cute; didn't read)
holtkam2 2 days ago |
redanddead 3 days ago |
frank_debug 3 days ago |
undefined 2 days ago |
erikig 2 days ago |