📚 studying evolution to advance scenius at scale 🌎 @atlasresear_ch 🎙️📝 #humansontheloop 🦚 ex @sfiscience @longnow @mozilla 🏅 grantee @cosmos_inst @osvllc
Santa Fe, NM
Joined October 2007
- Tweets23.3K
- Following3.3K
- Followers7.2K
- Likes24.1K
Pinned Tweet
Where to place your attention is without question a moral act. This is the understanding I am trying to help inject into the conversation around generative large language models.
The more people energize worst-case outcomes the more we guarantee them. If we can help people get enough distance from their knee-jerk fear response to reflect on their own minds and, as the bumper sticker says, not believe everything they think, we can interrupt the fear stream with an executive reallocation mandate that brings us rapidly and coherently closer to worlds we actually want to live in.
Silicon Valley is having a bad trip right now. It needs spiritual guidance if we are to live the promise of these tools.
Michael Garfield 💫 retweeted
a grant program quietly encodes a theory of reality and asks - what counts as knowledge, where novelty comes from, whether the future is predictable, whether people can be evaluated in advance, whether discovery can be planned, and how much uncertainty an institution is willing to tolerate
New episode on The Narrative #Brain:
@FritzBreithaupt and his lab looked at how stories evolve as they are retold between over 20,000 human subjects. And then compared it to the same stories changing over time as they are retold by ChatGPT.
Depending on who you are, what they discovered either clarifies the real difference between humans and AI, unsettles our ideas about selfhood by anchoring it in the rupture of experiential learning, or suggests a path forward for the future of technology.
Fascinating conversation with Fritz and cognitive neuroscientist @adamsafron for our world models mini-series!
Full episode and interactive discourse map for the whole Phil Trans A special issue anywhere you go for podcasts (but here)
Chapters
00:00 Intro
06:56 Meet Fritz Breithaupt
08:08 A Childhood Rupture
11:41 Narrative as Therapy
13:30 Telephone Games
18:26 Emotions as Story Anchors
21:04 Prediction and Multiverses
23:58 Unnamed Narrative Emotions
31:28 AI Retellings vs Humans
34:07 Perspective-Taking and Experience
43:58 Limited World Models
48:41 Curiosity and Scientific Creativity
54:03 Suspense and Surprise
55:26 Three Modes of Experience
01:02:40 Self Modeling Tradeoffs
01:07:56 Reinventing the Self
01:14:21 Genre Collapse and Play
01:19:53 Sharing Experience by Story
01:27:47 Continual Learning and Agency
01:32:42 Mortality and Selfhood
01:38:44 Self as Process Not Object
01:45:09 Plural Consciousness and Politics
01:48:08 Outro
#AI #cognitivescience #neuroscience #storytelling
Dig into the extensive show notes, transcript, and related discussions on world models, cog sci, and AI — upcoming episodes with @blamlab & @drmichaellevin:
humansontheloop.com/p/h-40
Michael Garfield 💫 retweeted
"actually enjoy your work as a primary principle" has never been more of a solid bet
with ai there is infinite degrees of freedom and directions you can move and work
grinding dampens imagination and constrains - complete opposite of what this new work calls for
🚨 Wow. WOW. I just witnessed the most involved and sophisticated phishing attempt I've ever seen and y'all need to know about this.
The TL;DR is that @beaconlayer reached out invite me onto their show, spent days going over the run-of-show in detail with clearly AI-assisted extensive knowledge of my work, booked the recording, and then tried to get me to install malware in order to join the call.
Some of you have probably already heard about the similar attempt made on @0xzak but this one was news to me. The difference is that this isn't just someone posing as a producer from a popular show, it's the actual show, which has episodes out and a sizeable audience on @X which makes them look legit. I suspect all of those other guests have malware installed already and I caution everyone to do a security sweep.
I've been cautioning people against deepfake scams for a decade so you'd think I'd have smelled something sooner — it takes real balls to go after cypherpunks specifically — but at least I caught it before they convinced another rube to commit a bash command and give up all of their passwords.
Here's the full story — I have the receipts, and now you do too:
1/13
At this point I am:
1) Glad I caught it, and somewhat impressed at how far they were willing to go to pull this off.
2) Worried that everyone else who presumably downloaded working vibe-coded streaming software from these goons has given up their passwords and recovery keys.
3) Absolutely warning as many people as I can, because even if they ditch all of these accounts and start from scratch the fact of it is that it is now trivially easy to steal or amass a ton of followers.
You can expect to see many, many more attacks like this in the months and years to come.
As if I needed any more evidence, this experience really drove home for me how important the work we're doing at @atlasresear_ch is: the Web needs intimacy gradients, needs sovereign decentralized reputation, needs security that can keep pace with the no-jurisdiction, near-zero-cost scams that are coming our way in the age of agent swarms.
I feel more bullish than ever on what my team is doing.
12/13
One last note:
The poetic timing of this elaborate phishing attempt is intense. Later today I'm hosting a salon with @krishnanrohit on Charles Stross' award-winning 2005 post-singularity sci-fi novel Accelerando, which is an absolute masterclass in AI-enabled scams.
I owe Stross a lot of credit for helping me think through the fine points of epistemic collapse and the technological arms races to come — Accelerando has its fingerprints all over my own speculative fiction. It got me out there ringing the alarm about voice clones nearly a decade before it hit cable news. And it's trippy to be living in a world where our insane future visions are starting to come true.
Don't believe me? Good. Critical thinking is going to serve you well:
humansontheloop.com/p/realit…
PS — If you want to join us for the call, hit me up! I promise not to force you to execute any CLI shenanigans to be a part of this conversation.
Actual human, actual podcast, I swear.
13/13
Michael Garfield 💫 retweeted
IP is definitely not dead, definitely overdue for a massive overhaul, and definitely will require more nuanced tracing.
The problem is that everyone seems to think that this will be fixed by arbitrarily starting that tracing at the moment that favors their exclusive claim on whatever they only managed to enclose thanks to the occlusion of their actual debts — to history, to the commons, to illegible labor, to myriad other subsidies.
God forbid AI steal your style but it's okay to claim the style is "yours" in the first place, like you're an island of original thought and not utterly embedded in a web of interbeing.
For the same reason information doesn't want to be free, there is no full-cost accounting — and any effort to enact a fully comprehensive back-paid royalty system for ideas is doomed to be partial, doomed to leave plenty of contributors high and dry, even as people rush into even more complete surveillance to stake their claims.
If the next equilibrium rewards "first legible actor" and starves everyone who made that work possible in the first place, we'll be no better than we are, and arguably worse.
But maybe the systems that will give us enough context to empower us to do genuinely novel work will be the same systems that finally disabuse us of this idea that anyone can make a permanent claim on anything. Period.
On the next economy (partial)
humansontheloop.com/p/sceniu…
On why IP needs a life cycle
humansontheloop.com/p/rent
