George Hotz, a developer active in the AI sector since 2014, reports that he currently experiences a 10x productivity boost using LLMs for programming. He cites his success in configuring Linux environments using GLM-5.2 to automate tasks like installing tmux. Despite his enthusiasm for LLM progress, video generation models, and coding agents, he rejects negative industry hype regarding the supposed risks of falling behind or the existential threat of AI superintelligence. He argues that the perception of models as world-altering entities is a strawman, noting that common fears surrounding AI-driven disasters resemble science fiction tropes from 1991 or 2016. Hotz contends that frontier AI labs struggle to capture value because AI progress is driven primarily by Moore's law and general computing advancements rather than proprietary efforts. He maintains that anti-open source narratives from these companies are motivated by a fear of commodification rather than legitimate safety or geopolitical concerns. Regarding software development, he notes that while he previously doubted the capability of models to program, he now views them as useful tools for productivity, comparable to compilers or regex generators. However, he warns that relying on current AI models can increase cognitive fatigue and produce low-quality content, or 'slop'. He concludes that AI is the natural continuation of the computer revolution he has admired throughout his career.

The author emphasizes that while current models offer productivity gains, they have yet to produce the wave of transformative software that such efficiency improvements might imply. He asserts that the valuation of large, closed-source frontier labs is inflated by marketing narratives designed to secure billions of dollars in funding by obscuring the commodified nature of the underlying technology.

George Hotz, a developer active in the AI sector since 2014, reports that he currently experiences a 10x productivity boost using LLMs for programming. He cites his success in configuring Linux environments using GLM-5.2 to automate tasks like installing tmux. Despite his enthusiasm for LLM progress, video generation models, and coding agents, he rejects negative industry hype regarding the supposed risks of falling behind or the existential threat of AI superintelligence. He argues that the perception of models as world-altering entities is a strawman, noting that common fears surrounding AI-driven disasters resemble science fiction tropes from 1991 or 2016. Hotz contends that frontier AI labs struggle to capture value because AI progress is driven primarily by Moore's law and general computing advancements rather than proprietary efforts. He maintains that anti-open source narratives from these companies are motivated by a fear of commodification rather than legitimate safety or geopolitical concerns. Regarding software development, he notes that while he previously doubted the capability of models to program, he now views them as useful tools for productivity, comparable to compilers or regex generators. However, he warns that relying on current AI models can increase cognitive fatigue and produce low-quality content, or 'slop'. He concludes that AI is the natural continuation of the computer revolution he has admired throughout his career.

The author emphasizes that while current models offer productivity gains, they have yet to produce the wave of transformative software that such efficiency improvements might imply. He asserts that the valuation of large, closed-source frontier labs is inflated by marketing narratives designed to secure billions of dollars in funding by obscuring the commodified nature of the underlying technology.