Moore’s law, named after Gordon E. Moore, the co-founder of Intel, states that the number of components on a single chip doubles every two years at minimal cost. It’s not exactly hard science, but his simple observation of how rapidly chip technology was progressing back in 1965, has proved to be pretty useful for predicting where technology will be in the years to come.
In his latest blog post, Sam Altman, CEO of OpenAI, makers of ChatGPT, just stated that ‘ Moore’s law changed the world at 2x every 18 months; this is unbelievably stronger. ‘ Altman was referring to the massive price drop we’ve seen for AI usage. For developers, this is measured in token cost, and the token cost to use AI has been falling about 10x every 12 months. Between early 2023 to mid-2024, the price per token for using ChatGPT has dropped about 150x.
AGI is coming
Technically, Altman got it wrong by writing “18 months” instead of “about two years”. The 18 months figure is a common misquote since there was a separate prediction by Moore’s colleague, Intel Executive David House that Moore’s Law would mean that chip performance would double every 18 months, with no increase in power consumption. So, a forgivable slip.
However, Altman’s point is that the falling cost of using AI is another indicator that artificial general intelligence (AGI) is “coming into view”. AGI is an artificial intelligence that is on a par with, or smarter than, human intelligence, and developing it is the whole reason that companies like OpenAI exist because the benefits of AGI could be truly world-changing, despite the dangers that we keep getting warned about.
It’s worth reading the whole of Sam Altman’s blog post because it very much reads like a warning that the future is coming sooner than we think and that we really start to need to prepare for how the world is going to change. Altman’s predictions are certainly mind-blowing. His three key predictions for the immediate future are:
- Scientific progress will likely be much faster than it is today
- The price of many goods will eventually fall dramatically
- The price of luxury goods and a few inherently limited resources like land may rise even more dramatically
Altman doesn’t predict these changes happening in 2025, but since he’s already gone on record saying that humanity will achieve AGI this year, the changes would appear to be right around the corner.
His message is also that we have some tough choices coming up regarding AGI and how freely it should be used, because one possible future, “we can see is AI being used by authoritarian governments to control their population through mass surveillance and loss of autonomy.”
In Altman’s vision, ensuring that the benefits of AGI are broadly distributed is critical to prevent this from happening.
He ends with a quite astonishing prediction: “Anyone in 2035 should be able to marshall the intellectual capacity equivalent to everyone in 2025; everyone should have access to unlimited genius to direct however they can imagine.”
It’s hard to know how to process the breadth of Altman’s vision right now. Companies like OpenAI are committed to developing AGI, but the dangers inherent in who controls this technology are something that Altman is urging us to start considering and talking about right now.
Leave a comment