RT @MichaelSurrago: @curtdoolittle So dope. Been using too. (Feels like psilocybin on coffee honestly.) Every fourth sentence when I teach…
Source date (UTC): 2021-10-08 00:54:12 UTC
Original post: https://twitter.com/i/web/status/1446277716488114176
RT @MichaelSurrago: @ThruTheHayes @curtdoolittle Also agree on the number.
my students have been introduced to it (kids version) last year…
Source date (UTC): 2021-10-08 00:54:01 UTC
Original post: https://twitter.com/i/web/status/1446277669088337924
It’s endlessly humorous that the BTC community cannot grasp it’s doing free R&D for the state. But it is. The end result will be something similar, but it will not replace state currencies because the competitive value of state debt can’t be eliminated in economic calculation.
Source date (UTC): 2021-10-07 16:11:18 UTC
Original post: https://twitter.com/i/web/status/1446146121307459590
Reply addressees: @wastelander113 @jordanbpeterson @dergigi
Replying to: https://twitter.com/i/web/status/1446145812153700358
IN REPLY TO:
Unknown author
@wastelander113 @jordanbpeterson @dergigi The new currency that fractionally divides and appreciates provides a vehicle for ‘savings’ while the primary currency that depreciates in value through increases provides a vehicle for state ‘debt’. This dual model maximizes the value of converting debt to production to savings.
Original post: https://x.com/i/web/status/1446145812153700358
My position is that we should maximize and optimize the biological form we have, and that we should construct machines we can cooperate with. And if possible create intelligent machines that are hyper moral if it becomes possible. “no harm to the body” would be my warning.
Source date (UTC): 2021-09-29 14:43:40 UTC
Original post: https://twitter.com/i/web/status/1443224968238600199
Reply addressees: @AdamNoEve666
Replying to: https://twitter.com/i/web/status/1443223620021596163
Ok. I would hug you for that if I could reach you. 😉
lol
Now take combinatorics. In other words computation just like language – because it is the same as language – can compute anything (is endless). Ie universal turning machines using categories rather than infinite memory.
Source date (UTC): 2021-09-29 14:33:38 UTC
Original post: https://twitter.com/i/web/status/1443222442588188675
Reply addressees: @HJoshington
Replying to: https://twitter.com/i/web/status/1443221961920888835
Quote of the day that made my day: in the context of quantum computing, at the quantum level “pretty much anything, of any material, will compute.”
Exactly. PRETTY MUCH EVERYTHING COMPUTES
Source date (UTC): 2021-09-29 14:31:56 UTC
Original post: https://twitter.com/i/web/status/1443222013812912136
And the reason I don’t fear AI as a deterministic harm is because I solved the problem of creating deterministically moral AIs. This doesn’t prevent people from creating harmful AIs – it just explains that it’s not necessary. And AIs can be even MORE moral than humans. https://twitter.com/curtdoolittle/status/1443204908002516996
Source date (UTC): 2021-09-29 13:41:49 UTC
Original post: https://twitter.com/i/web/status/1443209400999628809
https://twitter.com/curtdoolittle/status/1443204908002516996
I still can’t figure out if David Deutsch and I are saying the same thing because I can’t quite get my arms around his disambiguation between computation and quantum computation. I use computation vs calculation because I work in the opposite direction. I THINK that’s the same.
Source date (UTC): 2021-09-25 19:24:31 UTC
Original post: https://twitter.com/i/web/status/1441846093776658434
GPT3 is not novel it’s just cheap enough now, with cheap enough access to textual information, to produce ends. The only value in GPT3 is illustrating it’s just the number of neurons, the recursion (wayfinding) possible and the context (memory) that makes intelligence possible.
Source date (UTC): 2021-09-25 18:52:22 UTC
Original post: https://twitter.com/i/web/status/1441838003672924160
Reply addressees: @JaneSchwartz100 @Jonathan_Blow
Replying to: https://twitter.com/i/web/status/1441837090707099654
I understand that GPT3 is a ‘trick’ but is dumb as a rock. Which you’ll find if you use it at all. GPT’s memory (context) is about two pages of text. What context(equivalent) is necessary for demonstrated intelligence?
Source date (UTC): 2021-09-25 18:51:16 UTC
Original post: https://twitter.com/i/web/status/1441837727167696906
Reply addressees: @JaneSchwartz100 @Jonathan_Blow
Replying to: https://twitter.com/i/web/status/1441837090707099654