https://artificial-mind.blogspot.com/2025/12/the-excessive-spending-for-ai-bubble-is.html
""" The excessive spending* for "The AI Bubble" is not *a mistake*. It is not because "they don't understand the laws of scaling" etc. (see the comments to the Gary Marcus' post). This is [just] capitalism. (...) (...) Apparently the goal is to MAXIMIZE THE EXPENSES (which come from another source) and to cheat and deceive your "host" source that you optimize and that you spend the least possible (not enough), in order to take even more resources and to spend even more: this is [also a pattern of the] "consumer culture" and capitalism, not [of] "AI", LOL. (...) E.g. ML globally is paradoxically similar - its goal is presented as "optimizations", but how many times the same datasets are "relearned" from scratch on how many places and how many GPUs, how many useless benchmarks are testing things which are already known, improving the "SOTA" with 0.4 points, 4.2 points on MMLU, XYZA*. Thousands of papers [are] repeating almost the same, all countries [are] having "National AI strategies" and institutes (clones of the 2003 original with exponential financial requirements), yet the most of what's done everywhere is replication & multiplication with minimal variation and maximum spending - just like with the living organisms and their goals. (...) Globally nothing is "optimized". All is a trade-off - converting one currency to another one and cheating by accounting only a cherry-picked tiny fraction of the cost. * The broader and apparently paradoxical maximization of the inefficiency is explained in a more universal sense in the small monograph "Is Mortal Computation Required for the Creation of Thinking Machines?" at SIGI-2025* """ SIGI-2025 presented the huge biblical ~20 volume work "*The Prophets of the Thinking Machines: AGI & Transhumanism: History, Theory & Pioneers; Past, Present and Future", *nearly 5000 pages in total. The virtual conference SIGI-2026 has informally already opened on 1.1.2026 and will continue until 31.12.2026. It may possibly volumes or systems from the continuation: *"Genesis: Creating Thinking Machines" *and additional volumes as appendices to "The Prophets...". Corrections and updates are coming to the original Prophets as well. * The keyword is "excessive" - and not investment in the most efficient researchers which produce novel discoveries and ideas with zero resources or about zero (the so called "Singularity of Tosh"), but in structures which are "scaling" *the expenses more, *because this leads to profits/"optimization" at particular points of the system, which are measured. - see the volume "The first modern AI strategy was published by an 18-year old Bulgarian in 2003 and repeated and implemented by the whole world 15-20 years later (...)") on SIGI-2025. ------------------------------------------ Artificial General Intelligence List: AGI Permalink: https://agi.topicbox.com/groups/agi/Tc2616dc0a8577813-M5392b421a33e7301551ddbd5 Delivery options: https://agi.topicbox.com/groups/agi/subscription
