Victor Senkevich@VictorSenkevich
šøAGI Guru. CTO. šøAGI is closer than it appears in the mirror of our perception. šøThe problem is solvable if it is codable. Somewhere on planet Earth Joined February 2011-
Tweets1K
-
Followers58
-
Following32
-
Likes90
ā¢ The trivial truth is that real AGI is a reasoning AI (RAI) ā¢ Reasoning is a discrete / transactional sequential / recursive process of forming knowledge from data, using feedback to remove uncertainty ā¢ Reasoning is a mechanism for understanding x.com/VictorSenkevicā¦
ā¢ The trivial truth is that real AGI is a reasoning AI (RAI) ā¢ Reasoning is a discrete / transactional sequential / recursive process of forming knowledge from data, using feedback to remove uncertainty ā¢ Reasoning is a mechanism for understanding x.com/VictorSenkevicā¦
ā² There is no reality, only perception Ā©VS ā² Communication without perception is useless and meaningless š¬ Quote: ā¢ "The number Ļ and Bach's music exist only for those who perceive them" / š "Can Artificial Intelligence be Conscious?ā x.com/VictorSenkevicā¦
ā² There is no reality, only perception Ā©VS ā² Communication without perception is useless and meaningless š¬ Quote: ā¢ "The number Ļ and Bach's music exist only for those who perceive them" / š "Can Artificial Intelligence be Conscious?ā x.com/VictorSenkevicā¦
š¬ Google AI: "Accurate data + computing power = the key..." ā¢ No, definitely not ā¢ Verbal "accurate data" is an unattainable and unnecessary fetish ā¢ Betting on "computing power" means an erroneous "quantitative" approach to a "qualitative" problem x.com/VictorSenkevicā¦
š¬ Google AI: "Accurate data + computing power = the key..." ā¢ No, definitely not ā¢ Verbal "accurate data" is an unattainable and unnecessary fetish ā¢ Betting on "computing power" means an erroneous "quantitative" approach to a "qualitative" problem x.com/VictorSenkevicā¦
ā¢ Don Quixote fought against windmills considering them animate and hostile. Those who accuse LLMs of intolerance and bias are making the same mistake ā¢ LLM is just a statistical approximator ā¢ The best learning processes user's feedback in real time ā¢ Norbert Wiener knew it
ā¢ The AGI "Master Algorithm" is a small, recursive, and simple "protocell" / elementary von Neumann cellular automaton reproducing processes of thinking ā¢ An AGI instance in such an architecture is a recursively generated aggregation of such algorithms x.com/VictorSenkevicā¦
ā¢ The AGI "Master Algorithm" is a small, recursive, and simple "protocell" / elementary von Neumann cellular automaton reproducing processes of thinking ā¢ An AGI instance in such an architecture is a recursively generated aggregation of such algorithms x.com/VictorSenkevicā¦
ā¢ It is useless to "predict" the sequence of decimal places of the number Ļ ā¢ It is necessary to understand and explain the formula for calculating it ā¢ This is the whole difference between LLM and AGI x.com/VictorSenkevicā¦
ā¢ It is useless to "predict" the sequence of decimal places of the number Ļ ā¢ It is necessary to understand and explain the formula for calculating it ā¢ This is the whole difference between LLM and AGI x.com/VictorSenkevicā¦ https://t.co/XqdtWoHF2e
āØ Architecture is primary, resources are secondary ā¢ An ineffective architecture wastes resources x.com/VictorSenkevicā¦
āØ Architecture is primary, resources are secondary ā¢ An ineffective architecture wastes resources x.com/VictorSenkevicā¦
ā¢ In some cases, the lack of information can be compensated by sufficient motivation ā¢ In some cases, the lack of motivation cannot be compensated by sufficient information ā¢ "Motivation", but not "intelligence", is the keyword for subjectness / agency x.com/VictorSenkevicā¦
ā¢ In some cases, the lack of information can be compensated by sufficient motivation ā¢ In some cases, the lack of motivation cannot be compensated by sufficient information ā¢ "Motivation", but not "intelligence", is the keyword for subjectness / agency x.com/VictorSenkevicā¦
Demis Hassabis @demishassabis
354K Followers 122 Following Co-founder & CEO @GoogleDeepMind - working on AGI. Trying to understand the fundamental nature of reality. Also revolutionising drug discovery @IsomorphicLabsDonald Hoffman @donalddhoffman
39K Followers 24 Following Professor Emeritus of Cognitive Sciences at the University of California, Irvine. Research: Visual perception, evolutionary psychology, consciousness, AI.FranƧois Chollet @fchollet
465K Followers 767 Following Deep learning @google. Creator of Keras. Author of 'Deep Learning with Python'. Opinions are my own.IBM @IBM
719K Followers 4K Following We partner with developers, data scientists, CTOs and other creators to make the world work better.Google AI @GoogleAI
2.2M Followers 23 Following Google AI is focused on bringing the benefits of AI to everyone. In conducting and applying our research, we advance the state-of-the-art in many domains.Andrej Karpathy @karpathy
959K Followers 898 Following š§āš³. Previously Director of AI @ Tesla, founding team @ OpenAI, CS231n/PhD @ Stanford. I like to train large deep neural nets š§ š¤š„Thomas G. Dietterich @tdietterich
50K Followers 483 Following Distinguished Professor (Emeritus), Oregon State Univ.; Former President, Assoc. for the Adv. of Artificial Intelligence; Robust AI & Comput. SustainabilityxAI @xai
981K Followers 31 FollowingAndrew Ng @AndrewYNg
999K Followers 900 Following Co-Founder of Coursera; Stanford CS adjunct faculty. Former head of Baidu AI Group/Google Brain. #ai #machinelearning, #deeplearning #MOOCsNathan Benaich @nathanbenaich
50K Followers 31K Following solo member of investment staff @airstreet, brewing ambition @airstreetcafe, next token predictor @airstreetpressSubbarao Kambhampati ... @rao2z
16K Followers 29 Following AI researcher & teacher @SCAI_ASU. Works on Human-Aware AI. Former President of @RealAAAI; Chair of @AAAS Sec T. Here to tweach #AI. YouTube Ch: https://t.co/4beUPOmMW6Nassim Nicholas Taleb @nntaleb
1.0M Followers 2K Following Flaneur: probability (philosophy), probability (mathematics), probability (real life),Phoenician wine, deadlifts & dead languages. Greco-Levantine.Canaan. #RWRIPeter Wang š¦ @pwang
48K Followers 2K Following Chief AI & Co-founder @AnacondaInc; invented @pyscript_dev, @PyData @Bokeh @Datashader. Former physicist. A student of the human condition. bsky: @wang.socialLex Fridman @lexfridman
3.5M Followers 119 Following Host of Lex Fridman Podcast. Interested in robots and humans.Melanie Mitchell @MelMitchell1
43K Followers 647 Following Professor, Santa Fe Institute. More thoughts at https://t.co/nC43NHRozX.Cade Metz @CadeMetz
31K Followers 1K Following New York Times reporter, covering A.I., driverless cars, and other changes: [email protected]. My book, "Genius Makers": https://t.co/TJBqNRKR5Q.Pedro Domingos @pmddomingos
77K Followers 163 Following Professor of computer science at UW and author of 'The Master Algorithm'. Into machine learning, AI, data science, and anything that makes me curious.Paul Pallaghy @PaulPallaghy
160 Followers 42 Following AI engineer | Practical futurist | PhD physicist | AI & green energy enable everything | Systems engineer | Structural biologist | Pandemic epidemiologistDavid Chalmers @davidchalmers42
29K Followers 666 Following philosopher@NYU. consciousness, reality+, life, the universe, and everything.Paul Yacoubian @PaulYacoubian
43K Followers 2K Following Founder, CEO @copy_ai. Helping companies grow faster and more efficiently with the world's first GTM AI Platform. Angel investor (ai apps and infra)Irina Rish @irinarish
9K Followers 986 Following prof UdeM/Mila; Canada Excellence Research Chair; AAI Lab head https://t.co/UzlrC7ZrGF; INCITE project PI https://t.co/0rV7szd7rH; CSO https://t.co/XDhj6MEtUjShane Legg @ShaneLegg
51K Followers 57 Following Co-founder and Chief AGI Scientist, Google DeepMindDavid Deutsch @DavidDeutschOxf
118K Followers 62 Following Physicist. Author of The Fabric of Reality and The Beginning of InfinityBrendan Dolan-Gavitt @moyix
24K Followers 6K Following Associate Professor @ NYU Tandon. Security, RE, ML. PGP https://t.co/3WXr0RfRkv Founder of the MESS Lab: https://t.co/zGycrX3Gmn "an orc smiling into the camera" ā CLIPRob Toews @_RobToews
4K Followers 706 Following Partner @RadicalVCFund, AI columnist @Forbes. "the machine does not isolate man from the great problems of nature but plunges him more deeply into them."Gary Marcus @GaryMarcus
142K Followers 6K Following A leading expert on AI, spoke at US Senate AI Oversight Subcommittee. Founder/CEO of Geometric Intelligence (acq. by Uber), TED speaker, host Humans vs MachinesYann LeCun @ylecun
699K Followers 710 Following Professor at NYU. Chief AI Scientist at Meta. Researcher in AI, Machine Learning, Robotics, etc. ACM Turing Award Laureate.Joscha Bach @Plinz
128K Followers 746 Following FOLLOWS YOU. Artificial Intelligence, Cognitive Architectures, Computation. The goal is integrity, not conformity. https://t.co/rFUNzdYXuK@VictorSenkevich Yes thatās right. Probabilistic world models are the primary, grounded level of perception and understanding. The real one. I donāt understand what you mean by āpredictiveā prob. models, I make predictions w my primary grounded model all the time.
@VictorSenkevich @ddmatta Yeah Iāve become skeptical of this distinction. I guess this is where I side with eliminate materialism then.
@VictorSenkevich @ddmatta Iām not sure if Chalmers says it or not but Iāve heard people essentially argue that meaning depends on consciousness. Which would imply information requires consciousness. Or that information isnāt meaningful, which contradicts the definition of information.
@VictorSenkevich @ddmatta I would define meaning as Situatedness (the relationship between a node and the network it is embedded in). This allows us to give a precise definition of information as the signaling of Situatedness. Because of this, bio-chemical information is meaningful.
@VictorSenkevich @ddmatta Thanks for sharing your article. Interesting read!
SQLite is such a great tool for learning SQL and database design I wonder what's missing from that ecosystem that could help more people use it as a learning environment?
SQLite is such a great tool for learning SQL and database design I wonder what's missing from that ecosystem that could help more people use it as a learning environment?
āYou should never ask anyone for anything. Never- and especially from those who are more powerful than yourself. They will make the offer and they will give of their own accord.ā ā Mikhail Bulgakov The Master and Margarita
@VictorSenkevich @codegodzilla @ESYudkowsky @AndrewYNg (I will gladly volunteer myself as a negative sample when training AI girlfriends, promise š - though there might be a serious class imbalance problem here with finding any positive samples - my condolences š
@VictorSenkevich @codegodzilla @ESYudkowsky @AndrewYNg Awesome! So, let's develop AGI, so we all can finally drop out of the dating market. Win-win. Deal!
Wish solutions to your business challenges could be right at your fingertips? They can be with Government of Canada supports.
Announcing the xAI PromptIDE The xAI PromptIDE is an integrated development environment for prompt engineering and interpretability research. It accelerates prompt engineering through an SDK that allows implementing complex prompting techniques and rich analytics that visualizeā¦ x.com/i/web/status/1ā¦
@VictorSenkevich @Plinz AGI will be amazing. It will be able to tell us what AGI is.
@VictorSenkevich AGI is a technological system that is better at AI research than humans
Compelling case for the possibility of self aware AI, in other words AI with self direction (motivation). Methodically breaks down what it means to be a conscious agent and applies that to thinking sand.
Compelling case for the possibility of self aware AI, in other words AI with self direction (motivation). Methodically breaks down what it means to be a conscious agent and applies that to thinking sand.
@VictorSenkevich @pmddomingos It's called narcissism
We trained an AI using process supervision ā rewarding the thought process rather than the outcome ā to achieve new state-of-art in mathematical reasoning. Encouraging sign for alignment of advanced AIs: ā¦openai.com/research/improā¦
.@Microsoft-backed @OpenAI has released a new research paper with ideas on how to help prevent the #AI #hallucinations. ā¦@theCUBEā© ā¦@EvanKirstelā© ā¦@BillMewā© cnbc.com/2023/05/31/opeā¦
Breaking šØ OpenAI trained a new state-of-the-art AI model for mathematical problem solving through process supervision. Using this the AI can mitigate hallucinations and answer logically like humans. An AI's attempt to solve a mathematical problem step-by-step:
@VictorSenkevich @ciphergoth "We have to regulate motivation" is absollutely correct. "How can we regulate motivation" is the central question of AI safety, otherwise known as the alignment problem.