Tag
Chamath reveals the harsh reality of AI costs vs. returns: token costs double every 45 days, but downstream productivity gains are at most 5%. Large model capability improvement has hit an asymptote, and within the next 3-4 years, every company will face an ultimate reckoning between cost and benefit.
Andrew Chen argues that for common 'normie' prompts, frontier and open-source LLMs are nearly indistinguishable, implying pricing will trend to zero, open-weight models will dominate consumer use, and competitive advantages will shift to wrappers and bundling rather than raw model quality.
The paper discusses the small scaling exponents of large language models, arguing that they indicate an unsustainable regime in terms of energy resources. It also examines the 'pedestal effect' and draws analogies with fluid turbulence to comment on data smoothness.
Saagar Pateder analyzes the diminishing marginal returns of AI intelligence for consumer and enterprise tasks, and predicts that open-weight models will diffuse globally by 2029, based on historical trends in model performance and cost.
A distinguished engineer at a hyperscaler argues that AI models are hitting diminishing returns in software engineering tasks, as he finds little difference between Claude's Fable 5 and previous Opus models, and predicts local models will soon provide comparable value.
A reflection on how the excitement around new AI model releases has faded compared to the early days, drawing parallels to annual smartphone launches.