Sam Altman says OpenAI wants to sell intelligence like a utility
During a recent appearance at BlackRock in Washington, D.C., OpenAI’s Sam Altman, shared his vision for the future of AI. At one point saying, “We see a future where intelligence is a utility, like electricity or water, and people buy it from us on a meter.”
Altman was describing a world where AI becomes a foundational infrastructure, something woven into everyday life so deeply that consumers and businesses simply “plug into” it the same way they rely on electricity, Wi-Fi or running water.
One thing I like to bring up:
If current LLMs are so intelligent, can they write their own inference code?
…Or even make modest code contributions to those bases?
The answer is a resounding “no.” See: llama.cpp’s contribution policy, vllm’s PR history, janky diffusion UIs and such.
The severe bottleneck to efficient inference (much less training) of a variety of models is good code contributors, and AI is doing little to help. And, ironically, if it ever was smart enough to do that, ChatGPT the service would collapse in like a week because a ton of backlogged integration would get done, and the open source/open weights space would explode.