Piper Welch retweeted
.@KamBielawski just showed @ALifeConf that evolving things with "insides" which can evolutionarily change, neutrally, creates plateaus along which evolution can walk to escape from local optima.
bit.ly/4cQcLqe
Photo credit: @HirokiSayama. #ALIFE2024
Piper Welch retweeted
Hot off the presses at Chaos -
"Evolving higher-order synergies reveals a trade-off between stability and information-integration capacity in complex systems" w/ @DoctorJosh 1/n
pubs.aip.org/aip/cha/article…
Piper Welch retweeted
Looking to hire a postdoc to do AI-designed biology with @drmichaellevin and myself.
Please RT; DM if interested.
Here's what happened the last time we teamed up:
youtube.com/watch?v=8WyWFAS9…
Piper Welch retweeted
Really excited about this paper, featuring brand new PhD @GizemGumuskaya 🎉, collaborator @sjmgarnier, & undergrad students P. Srivastava, B. Cooper, H. Lesser, B. Semegran:
onlinelibrary.wiley.com/doi/…
What I personally think it means: thoughtforms.life/meet-the-a…
Piper Welch retweeted
#xenobots were made from frog cells (Xenopus laevis).
Now introducing "anthrobots".
Guess what they're made from.
nature.com/articles/d41586-0…
Piper Welch retweeted
AI can now generate images, whole movies, proteins, xenobots...
...how about mathematical descriptions of the world?
Lab member Amanda Bertschinger presented on why "neural symbolic regression" is so difficult at @ECMLPKDD yesterday.
Well done Amanda!
youtu.be/0sLt5hW0-j0
Piper Welch retweeted
Replying to @NewYorker
@NewYorker article out today about the great work of my colleague Mark Blumberg.
Featuring cameo appearances from myself and our "dreaming" robot, the Evil Starfish (past work w/ Victor Zykov and @hodlipson).
newyorker.com/science/elemen…
or
jbongard.github.io/docs/What…
Piper Welch retweeted
Susan Stepney: "Is ChatGPT conscious?"
Anil Seth: "No."
Ted Chiang: "No."
(Applause)
#ALIFE2023 @ALifeConf
Piper Welch retweeted
#xenobots are AI-designed organisms.
But what if next-gen biobots can't be designed by AI?
Introducing ADS's: AI-designed swarms!
@PiperRWelch will be presenting this, her new work, at #ALife2023 on Tuesday.
Joint work w/ @csgrasso10, @GizemGumuskaya and @DrMichaelLevin.
Piper Welch retweeted
Replying to @PiperRWelch
@PiperRWelch -- new lab member, @NSF GRFP recipient, and #xenobot team member -- was just profiled by our university.
She'll be presenting her work at the #ALIFE2023 conference in Japan this week. If you're there, check it out!
uvm.edu/news/ovpr/redesignin…
Piper Welch retweeted
Think you know what intelligence is?
Computation?
AI?
Embodied cognition?
Think again.
Take this red pill, and let @ParsaAtoosa , presenting Tuesday at @GECCOConf, show you how deep the rabbit hole of embodiment really goes...
Piper Welch retweeted
AI must understand the world to act safely within it.
To that end, @PigozziFederico will describe a robot at #GECCO2023 this week that uses its body to learn about the world in a new way:
Each body part "votes" about what is/isn't possible, at all times.
youtu.be/IGYQAzYiJvE
Piper Welch retweeted
How can #robot swarms work together safely?
@csgrasso10 has found that they need to push against their neighbors in diverse ways, and get pushed back by them in reliable ways, *quickly*.
Come see her describe this at @GeccoConf today.
And vote for this, a best paper contender!
Piper Welch retweeted
With the publication of several of the Xenobot papers, a lot of ideas were brought up that were too hard to properly discuss in short format. @DougBlackiston, @Kriegmerica, @DoctorJosh, and I put our thoughts on the most interesting open issues here:
liebertpub.com/doi/full/10.1…
Piper Welch retweeted
Open World Chat! Geek science episode 2! @PiperRWelch and @kamtb28 talk cutting-edge AI and biobots with @jim_dandeneau while I kind of drool glazy-eyed and scratch my ass because science is hard. #AI #biobots #Apocalypse openworldchat.com/episodes/s…
Piper Welch retweeted
Embodied AI, and @uvmvermont, on the front page of the @nytimes today.
nytimes.com/2023/04/11/scien…
















