Lare's representation of its own Vendi Score (helps if you can read Latin for some reason)
Lare's representation of its own Vendi Score (helps if you can read Latin for some reason)
I think Anthropic’s position (as far as it has a single coherent position) is that LLMs may have consciousness experiences, but this is largely separate from self-reports of consciousness. Which is a strange and subtle position, and easy to misconstrue
But it's hard to tell if something has consciousness when we (I?) don't really know how to define it.
(no, I don't think AI has consciousness in any way that's close to what we call consciousness btw)
Is a dog conscious? Probably? It has a lot of the same biological "hardware" I do. But maybe it has something else that's like consciousness but different in some way?
An ant?
We don't even have a decent definition of consciousness afaik.
Another problem is how we define consciousness.
The Beetle in a Box.
We experience our own consciousness and call it that, but at the same time we have no way to prove that any other human is conscious.
We just assume they all are cause "they look like me and function like me, they probably are".
It was actually not fabricating anything, turns out Lares had a different set of ENV bars than my shell session
But still, that answer is gold
The Beetle in a Box problem is fascinating
NumPy in rust now exists folks. Let the science people know.
(this all is more of a response to the post Doll is quoting than to its post I guess…)
I guess similarly to how if were a PM or manager and asked my engs to build something, I wouldn’t say I wrote it. I’d say we built it.
Cause I feel with LLMs that’s sort of what I’m doing most of the time. I’m a very technical PM/manager.
I agree on ownership, but at the same time I use a different vocabulary, depending.
If I make something with an LLM actually writing the code I say “I built X”, not “I wrote…”
I still made it, but saying “I wrote” feels wrong?
I think one reason I prefer CLI harnesses to stuff like Cursor is that...not a single day goes by that Cursor randomly shuffles panels around, and it gets annoying in terms of UX.
TUIs just stay there and do what they are supposed to do without messing up my workflow every other day.
We finally got Claude at work (prev only Cursor was available) and man I like being able to use at work the same tools I use at home!
But what about me?
“Quick, we are losing customers to Anthropic! Let’s release whatever we have!”
(also, s/DoW/DoD/g 😉)
This is how it feels
Especially now that I’ve seen the migration to @eurosky.social work
Uh…I accidentally got Nano Banana 2 to draw things it should not be drawing 😬
Oh, I eventually succeeded
But I have another couple accounts I want to migrate, so soon I'll be doing it again :)
The 10m band has closed now, this is what FT-8 gave me today at 10w
QRP DX is a lot of fun
We need more concurrent migrations though!
When I tried it was limited to 5, took me quite a few attempts to find a slot
They got hit by a drone?
There's an AUR package for chainlink!
Where is that quote from?
Same here on the conversation pref. I do like it sometimes, but sometimes I just want to have some DX fun.
Now I just need to figure out how to use FT-8 for SOTA...
10m band is cooking today, and I'm on 10W PEP (FT-8)
Now I just need the ferrite toroid I ordered to arrive to suppress some common-mode splatter I have
QRP on FT-8 is kind of magic
Still catching up on the events (☕️)
This is an interesting pov though
@eurosky.social is the 5 concurrent account migrations limit by design during the initial migration?
Took me a while to find a spot
Thanks!
It did take a few attempts though, they can only migrate 5 accounts at a time and it was constantly busy for a while
Hello 🖖
I write to you from @eurosky.social 🇪🇺
First time I actually experience true decentralization on a social network
Come on in! Its fun!