#lisp
#lispyGopherClimate #live #podcast at 000UTC Wednesday on https://anonradio.net:8443/anonradio by @SDF. (Tuesday night in either America).
It looks like we'll have @masinter about #medley #interlisp https://online.interlisp.org/user/login https://interlisp.org/software/ "bringing this historical software to life".
New @interlisp account! @amoroso
https://larrymasinter.net/ https://larrymasinter.net/85-ai-tutorial.pdf #GOFAI #AI
#live in #lambdaMOO as usual! <- Participation <3
#retrofuturistic #computing #unix_surrealism
Let everyone know!
https://lispy-gopher-show.itch.io/lispmoo2/devlog/835759/repeatedly-eval-qt-good-old-fashioned-ai-in-my-lispmoo2
#GOFAI #itch_io #devlog #gamedev #lispgames #commonlisp #moo
I added a puppet prototype for my own #lispmoo2 submission to my own #lispmoo2jam.
Puppets are actually crazy. You just pass her a piece of paper and she does whatever's on it, up to and including possessing you and driving you around if that was an implication of one of the commands you gave her that you didn't really think through.
Having computer players working with re/writeable-scripts-as-objects ++
If you like estimates of which #technology R&D approaches are likely to succeed, or interested in where #AI is headed, you might enjoy this Metaculus question. Open for comment now & prediction in 2 days. Resolves in Dec.
"Will conventional pre-#LLM #GOFAI (graph search, tree search, game play or symbolic logic) be a part of the top-ranked AI in the 2024 Abstraction & #Reasoning Corpus (#ARC) competition?"
#Metaculus is a nonmonetary (#karma-based) #PredictionMarket.
Out now: Klaus Mainzer (Ed.) (2024): Philosophisches Handbuch Künstliche Intelligenz. Wiesbaden: Springer Nature.
I contributed a chapter about #ELIZA and its sources in analytical philosophy of speech.
Last friday the @Weizenbaum_Institut published their website about Jospeh #Weizenbaum and his work. I contributed a chapter about the #ELIZA #chatbot
@Jacob @pluralistic @HalvarFlake
HT x sharing, enjoyed listening:
• Compare current to #gofAI begging #ML algos & tools
• Cool #RichSutton #TheBitterLessonInML remark that generic algos, w/o use of larger real datasets for training models, will beat expert algos & shorter datasets.
• Remark of synthetic data for training, yet employ with caution, and only helpful for some cases.
• Cool Q&A long answer for now is too late for keeping free internet datasets non-contaminated of synthetic BS .
weekend yt lecture suggestion:
“What's wrong with #LLMs
&
what we should be building instead”
ಠ_ಠ
issues:
• Incorrect & contradictory answers
• Dangerous & socially unacceptable answers
• Expensive to train & lack of updatability
• Lack of attribution & poor non-linguistic knowledge
challenges:
• Sepparate Linguistics from World Knowdledge
• Functioning Episodic Memory
By Veteran since #GOFAI times, #MLPioneer #JMLR Journo etc:
Prof #ThomasGDietterich @tdietterich https://youtu.be/cEyHsMzbZBs
The author is author well-educated regarding AI because he mentions chatbots, not LLM or AI ML. Go #GOFAI
Mind blown by Paul Lessard's explanation of how #CategoryTheory applies to #DeepLearning, on #MLST (#MachineLearning Street Talk podcast). The extra level of abstraction of CT gives you a "Lego set for the universe" (and #AGI).
Glad the #EA folks are still thinking about this stuff and starting to be a bit more sceptical about #LLMs, recognizing that LLMs need to be augmented with #GOFAI, like https://oval.cs.stanford.edu & https://symbolica.ai
There is more than LLMs, transformers and generative AI. Sometimes the discussion about AI seems to be very narrow. Don't forget there are many other machine learning technologies available like neural nets, decision trees, SVM etc. I especially like Lecun's V-JEPA as far as i can understand it.
Personally i think the way ahead lies in combining various approaches, which is already happening i think.
https://ai.meta.com/blog/v-jepa-yann-lecun-ai-model-video-joint-embedding-predictive-architecture/
#AI #LLM #generativeAI #GOFAI #V-JEPA #Lecun