“The cost of AI is falling at a rapid pace, while AI models
are catching up with human skills at an ever-increasing rate. OpenAI
researchers, the creators of ChatGPT, have just released some surprising
statistics.
OpenAI, a company led by Sam Altman, is one of the main
beneficiaries of the rapid development of AI
According to the latest OpenAI
report, the cost of using a given level of AI capabilities is falling rapidly –
every 12 months it drops… about 10-fold.
The researchers claim that lower prices lead
to much greater use. Suffice it to say that OpenAI AI is used by half a billion
people worldwide, including 3 million developers.
How fast is AI developing and getting cheaper?
The effect of “cheaper AI” was
observed in the example of the change in the cost of a token between GPT-4 in
early 2023 and GPT-4o in mid-2024. “The price per token during this period fell
by about 150 times” – OpenAI experts write.
At the same time, researchers from the company led by Sam
Altman indicate that the amount of time needed to improve an artificial
intelligence model is constantly decreasing. In other words, AI models are
catching up with human intelligence at an increasingly rapid pace.
"The typical time it takes a
computer to beat humans in a given comparative test has dropped from 20 years
after the test was introduced to five years," we read in the OpenAI
report.
But this data is also outdated,
because it refers to statistics from last year. Today, this indicator is
already less than two years.
And that's not the end. "We see no reason why this
progress should stop in the near future," the creators of ChatGPT
emphasize.
Europe is backward in AI
For now, it is mainly the Americans and the Chinese who
benefit from this boom. On the Old Continent, the development of AI is clearly
lagging behind.
According to the latest Eurostat
statistics for 2024, only 13.5 percent of European companies use AI technology.
This figure stands in stark contrast to the EU Digital
Decade’s goal of achieving a 75% AI adoption rate among enterprises by 2030.
At a deeper level, there is a huge
gap between larger organizations, which are showing higher adoption rates
(41%), and small and medium-sized businesses, which are lagging behind (11%).
“Currently, AI adoption in the public sector lags behind the
private sector. A European Commission study found that complex public
procurement procedures and concerns about bias in AI-based decision-making are
significant barriers to wider AI adoption in European public institutions,”
OpenAI points out in its report. It notes that addressing these issues and
encouraging wider AI adoption would deliver “more effective and efficient
public services, saving taxpayers money and freeing up resources for increased
investment in key areas".”
Komentarų nėra:
Rašyti komentarą