Special AI Meetup feat Joscha Bach & Rachel St. Clair - are LLMs enough?

RSVPs on AICamp have now closed due to Microsoft policy.

Please join the waitlist on AICamp and RSVP here. Note RSVP here does not guarantee entry at the door!!

We will try to provide an email update to those who have RSVP'd here around 5pm on Tuesday on the status of things.

Join us for another joint AICamp and Boston Astral Codex Ten event.

Sponsored by Boston ACX, Simuli, and Microsoft.

Agenda for the evening:
5:00 - 5:30 pm - setup
5:30 - 6:20 pm - socializing with pizza, veggies, and drinks
6:20 - 6:30 pm - intro to AICamp, ACX, and the Boston AI scene
6:30 - 6:50 pm - talk by Rachel Aileen StClair (Simuli)
6:50 - 7:05 pm - talk by Peter Sutor Jr (UMD)
7:05 - 7:15 pm - short break
7:15 - 8:15 pm - talk by Joscha Bach (Liquid AI), followed by discussion
8:15 - 8:30 pm - wind down

Speaker Bio:
Joscha Bach is an iconoclastic research working in AI and cognitive science. He previously was a researcher at the MIT Media Lab and at Harvard. Between 2021 - January 2023 he was Principle AI Engineer at Intel's Cognitive Computing group. He currently serves as AI strategist at Liquid AI and on the advisory council of the AI Foundation. His website is [http://bach.ai](https://l.facebook.com/l.php?u=http%3A%2F%2Fbach.ai%2F%3Ffbclid%3DIwAR0snv_Fw3TQbtihlFIypn5VF98_5U-NvCa6zwZEO30n-v9S2BAy2O24Bs4_aem_Afb0ZcWwfZSTut9h010F_s9FMUYoSvy4Eb_EBOJLa357zFpk9qMkiL4PFJ6cl7Esok0BDj3LypKGZQrQOPb2Iq3l&h=AT12PyrVG9Hnf-fO1egGe9Iy9UzTpUI_eMn7_s524KpigAwDjH6UGhxUQF9m6KCMiH_K64Mf9HO3oZgKuuk_hwCNg5O-CccqOvEgHBLgoPK-c3ZpZet1c0Uoa3rSwgPnGD9_rjpsARolFAkZnf7qrDg&tn=q&c%5B0%5D=AT04qmrsGWk2htdTPGqS912MiMZkCUXcAtlBsKF3yn_s6IvJKikEjEDoBTyJ2DQyUuRZi8kQozNM5FknEkQsJjp9_ehlJuKBGXZp84T43mphytNaB-NAVTZCMNhRYE9aPlI) and his Twitter/X handle iz @Plinz.

Talk Title & Description:
Is the LLM the last invention humanity needs to make?
The success of LLMs is taking the world by storm. The Turing test seems to have been obliterated years ago, and more applications and use cases are unlocked every week. Yet many of us are skeptical: are LLMs sufficient to carry us all the way to AGI? Are they the way to creating superhuman intelligence, a transitional development, or a dead end?

Speaker Bio:
Rachel St. Clair, Ph.D. is founder and CEO of Simuli (https://www.simuli.ai). She did her Ph.D. in Brain Sciences with Susan Schneider at Florida Atlanta University. After her Ph.D. she became a research fellow at the Center for Future Mind. She has published over 20 papers on AI and cognitive science. Rachel's company, Simuli, makes processors that use advanced math and an innovative hardware layout to process more information per transistor by compressing information before processing. Rachel envisions an AGI as a computer system which can outperform human abilities in nearly all tasks while helping humanity avoid existential risks.

Talk title & description:
Scalable AI -- Lessons from Biological Architectures for Cognitive Ones
Resource management is a tool nature has crafted over years of evolution. This tool not only has helped organisms and humans alike scale, but grow intelligence inside complex systems. Our focus will be on the two driving mechanisms of resource management as it relates to intelligence -- how to learn new things without forgetting old things and how to bootstrap a-priori knowledge. We'll discuss how these mechanisms work, their relation to existing AI. Finally, we'll showcase some recent work in our new cognitive architecture "ReArc".

Speaker bio
Peter Sutor is a Ph.D. student at the University of Maryland in the Department of Computer Science. His work focuses on vector encoding. He currenty works as an engineer at Simuli. He has published over 14 publications in high-dimensional computing and served as conference chair for the Hyperdimensional Computing Conference (HDC).

Talk title and abstact:
Efficient and Robust Vision Learning with Hyperdimensional Computing
Hyperdimensional computing allows low memory encoding of data that can be efficiently computed in hypergeometric spaces. These spaces afford few-shot learning techniques without suffering accuracy when compared to traditional deep learning methods. We'll show a new method of AI for classification that is cheaper to train and use.

RSVP on Less Wrong
Check out the Boston in-person events listing.