Expressing his views on the economics of AI, Altman also said it is critical to ensure that the benefits of AGI are broadly distributed.
OpenAI CEO Sam Altman said in a blog post on Sunday, February 9, that continuous and predictable gains in AI performance can still come from spending “arbitrary amounts of money.”
“The intelligence of an AI model roughly equals the log of the resources used to train and run it. These resources are chiefly training compute, data, and inference compute,” he said, adding that there was “no reason for exponentially increasing investment to stop in the near future.”
However, he also believes that the cost for users to access AI systems at a given level will fall by 10x every 12 months, which will lead to much more use of these AI systems. “You can see this in the token cost from GPT-4 in early 2023 to GPT-4o in mid-2024, where the price per token dropped about 150x in that time period,” Altman said.
Altman’s personal essay, highlighting his observations on the economics of AI, was posted on the eve of the highly anticipated, two-day Paris AI Action Summit which kicked off on Monday, February 10. He is reportedly expected to give a talk on the sidelines of the Summit on Tuesday.
His remarks follow the disruptive emergence of DeepSeek, a Chinese AI startup that claims to have developed a cutting-edge AI model at a small fraction of the cost of US rivals. The announcement last month had rattled global markets and sent shock waves across the tech industry.
OpenAI is reportedly in talks to raise up to $40 billion in a funding round. The startup behind ChatGPT as well as other corporate partners such as Microsoft, Nvidia, Arm, Oracle, and Softbank have pledged $500 billion to build out AI infrastructure such as data centers, energy plants, power lines, and more in the US over the next four years.
What else has Altman said?
On the debate between open-source AI models and closed systems, Altman said, “Many of us expect to need to give people more control over the technology than we have historically, including open-sourcing more, and accept that there is a balance between safety and individual empowerment that will require trade-offs.”
A few days ago, Altman had said that OpenAI needed to “figure out a different open source strategy” amid the growing popularity of DeepSeek’s open-source, reasoning model R1.
In Sunday’s blog post, Altman said that it is critical to ensure that the benefits of artificial general intelligence (AGI) are broadly distributed. “In particular, it does seem like the balance of power between capital and labor could easily get messed up, and this may require early intervention,” he said.
A weakly defined term, AGI generally refers to a system that can solve complex problems across various domains at a level that equals or surpasses human intellect.
While Altman continues to believe that AGI-level AI systems are near, some experts have argued that it will take another decade or so to get there. The OpenAI CEO envisions AGI to be deployed on a massive scale, with millions of AI agents “in every field of knowledge work.”
“[AGI systems] will not have the biggest new ideas, and it will be great at some things but surprisingly bad at others,” Altman wrote.
On the economic impact of AGI, Altman opined, “The price of many goods will eventually fall dramatically (right now, the cost of intelligence and the cost of energy constrain a lot of things), and the price of luxury goods and a few inherently limited resources like land may rise even more dramatically.”
He also hinted at “some major decisions and limitations related to AGI safety that will be unpopular” and warned of AI being used for mass surveillance by authoritarian governments.
In a footnote, Altman clarified that his thoughts on AGI do not have any bearing on OpenAI’s relationship with Microsoft.
The two tech companies reportedly have an internal definition of AGI as an AI system that generates at least $100 billion in profits. Over the years, Microsoft has invested more than $13 billion in OpenAI. “We fully expect to be partnered with Microsoft for the long term,” Altman wrote.