"Following two years of immense hype in 2023 and 2024, this year felt more like a settling-in period for the LLM-based token prediction industry. After more than two years of public fretting over AI models as future threats to human civilization or the seedlings of future gods, it's starting to look like hype is giving way to pragmatism: Today's AI can be very useful, but it's also clearly imperfect and prone to mistakes."
"There's a lot of money (and rhetoric) betting on a stratospheric, world-rocking trajectory for AI. But the "when" keeps getting pushed back, and that's because nearly everyone agrees that more significant technical breakthroughs are required. The original, lofty claims that we're on the verge of artificial general intelligence (AGI) or superintelligence (ASI) have not disappeared. Still, there's a growing awareness that such proclaimations are perhaps best viewed as venture capital marketing."
After intense hype in 2023–2024, AI shifted toward pragmatism in 2025 as models proved useful but error-prone. Expectations of imminent AGI or superintelligence were delayed amid consensus that further technical breakthroughs are needed. Commercial model builders prioritized sellable, reliable AI-powered solutions to generate revenue now. The year included sharp contrasts: high-level claims of AGI capability alongside modest model improvements such as GPT-5.1 learning punctuation behavior; enormous market valuations for companies like Nvidia amid warnings of an AI bubble; and plans for vast data-center energy consumption that underscore infrastructure and sustainability debates.
Read at Ars Technica
Unable to calculate read time
Collection
[
|
...
]