Has the #Artificial Intelligence industry officially shifted from a commercial, democratic ecosystem into a tightly controlled, national security apparatus reminiscent of the Manhattan Project? #ScholER
Alternative view - the LLM Gen AI industry is dying of lack of usage cases to justify the huge expenditures; the most compelling usage being malware/security threats is not a good look; otherwise, super-humanly fast coding is akin to the Concord super sonic travel - an expensive and dangerous solution mismatched to real world problems and benefits.
The chance that true ASI or even AGI emerges from the LLM Gen AI paradigm is nil - even with infinite scaling, self-improving algorithms, etc. - due to one simple reason: given the best of human generated general purpose training data is largely wrong or wrongly understood [reference “Most Published Research Findings Are False” John Ioannidis, PLoS Med 2(8): e124 (2005)] there is zero possibility to algorithmically make sense of that body of knowledge better than human experts in the fields with domain and context knowledge can. End of story.
If and when AI directed robots are collecting data, that may change, but it remains an open epistemological question, considering the unfathomable complexity and intractability of our physical world (where we can’t grok the 3 body problem in Newtonian physics, particle-wave duality in quantum physics or dark matter or energy in cosmology, etc..) let alone more complex biological, ecological, social, political domains.
Unfortunately, all this talk of AGI/ASI obscures a very real and immediate danger: purposefully or accidentally mis-aligned agentic Gen AI, with super-human coding/malware skills, could be a serious if not existential threat to our fragile, network dependent society in the near future. If “The unavoidable price of reliability is simplicity.” C.A.R. (Tony) Hoare’s Turing Award lecture, The Emperor’s Old Clothes (1980) – then the unavoidable price of extreme complexity (trillion+ parameter agentic Gen AI that is not understandable) may be extreme fragility; the sooner this is recognized, the better chance there may be to mitigate the proverbial unintended and unwanted consequences.
https://albertoromgar.med..
