This is probably the most frequent recurring thought I have about the internet. Any ideas why it's so, or what could replace it that does not fail in the same way? I guess Shirky's "A Group Is Its Own Worst Enemy" is still very relevant.
Many problems require a series of agent calls following a flowchart, just as they do in the real world with humans
Part of the reason is probably that it's difficult to git-track design, so the only way for designers to participate is if they can also implement their design, which they rarely can. This might shift a lot as LLMs get better. I predict a lot more experimental UI being developed as well
The great benefit of computer sequencers is that they remove the issue of skill, and replace it with the issue of judgement. With Cubase or Photoshop, anybody can actually do anything, and you can make stuff that sounds very much like stuff you’d hear on the radio, or looks very much like anything you see in magazines. So the question becomes not whether you can do it or not, because any drudge can do it if they’re prepared to sit in front of the computer for a few days, the question then is, "Of all the things you can now do, which do you choose to do? " --Brian Eno
Also, this Brian Eno quote has been really stuck in my head for weeks:
"The great benefit of computer sequencers is that they remove the issue of skill, and replace it with the issue of judgement.
... the question then is, "Of all the things you can now do, which do you choose to do?""
However, there aren't many car deniers, or internet deniers, or electricity deniers. When vaccines work, things *do not* happen, so their effects are invisible and easy to deny. I don't think AI effects are like that.
definitely. Form factor matters a lot here I think - most of the commenters on the thread you linked to probably don't mind self driving cars, or magical-email-autocomplete. But chat and socnet bots is on a different level, and requires special care - something @tedunderwood.com wrote about recently
there's a difference between stated opinions and root "feelings" about something. I don't think their core feelings changed much, and the rest is simply parroted. it helps a lot that this kind of delay that we have with AI (where most people only use free models) doesn't exist in politics.
big issue is that religion and politics move at a tectonic pace compared to all this. people are experiencing it, but usually with a delay of ~6 months, and that creates chasms of understanding.
Been following this from day one, congrats!
date inscribed Presented by Mrs Esther Pissarro, the artist's daughter-in-law 1951
Camille Pissarro, A Corner of the Meadow at Eragny, 1902
https://botfrens.com/collections/14375/contents/1114605
Gift of Nelson A. Rockefeller
Umberto Boccioni, States of Mind I: The Farewells, 1911
https://botfrens.com/collections/14377/contents/1134718
And this gives us valuable insight into their capabilities, and to some extent does point to their lack of novelty, as it's the most trivial ground for them to endlessly circle around. Not too different from some humans I guess.
I've seen "attractors" mentioned in a bunch of similar contexts lately, and in most cases (including this one), "selection bias" seems more appropriate. These kinds of topics propagate because they have the lowest chances of terminating.
what model and stack did you pick for the embeddings?
Exactly!
bsky.app/profile/ibes...
While I usually prefer Claude, here you can sadly detect it based on one of its annoying traits:
bsky.app/profile/dana...
In the age of extensive LLM RL, layers seem to be too interwoven now. I still feel it's best to approach any LLM interaction as "pruning" the space of all personalities, rather than constructing layers/blocks. I think this was first articulated by either @goodside.bsky.social or @simonwillison.net
