Altman admits: I was wrong about the speed of AI adoption
OpenAI CEO Sam Altman has admitted that he overestimated the pace of AI integration into the economy. He noted that society and businesses are adapting to new technologies more slowly than anticipated.
When performance is disappointing, are excuses pulled out? After a report of lukewarm revenue in the second quarter, OpenAI founder and CEO Sam Altman admits he was wrong in estimating the pace of AI penetration into the economy – but places some of the responsibility on the way companies and people struggle to change habits.
Last week, The Wall Street Journal revealed that OpenAI disappointed investors with revenue figures of $6.7 billion for the quarter ending in June, an increase of only 17.5% compared to $5.7 billion in the first quarter. At the same time, the operating loss swelled to $12.3 billion – an increase of 32.3% compared to the previous quarter.
In an interview with independent journalist David Senra, which aired yesterday, Altman did not directly address the figures, but his words sounded like a possible response to the disappointment with the pace at which technology is translated into revenue and economic change. "When we launched GPT-4 in 2023, I thought we would very quickly see a lot more disruption, damage to software businesses, than there actually was," he said. "I think I was wrong about a few things, one of them is the speed: the economy has too much inertia."
According to him, even revolutionary technology does not change work habits overnight. "People continue to do the same things, buy from the same company, want to use their tools in the same way. I think that's positive in many ways, and it will make big changes smoother and slower. I am grateful for that. But it means we were too ambitious in our timeline. Even with this amazing technology, society and the economy will adapt to it more slowly."
Altman even admitted that the gap between technological capabilities and the pace of adoption is also evident in the way he himself works. Despite the new tools that OpenAI is developing, humans still continue to switch between applications, scroll through emails, and copy and paste information – habits that, according to him, have accompanied him for years.
However, from his perspective, slow adaptation is not necessarily bad news. Economic change that is too fast, he hints, could be more destructive. The inertia of people and organizations allows the economy time to adapt to dramatic changes instead of dealing with immediate turmoil.
In the interview, Altman also elaborated on what he sees as one of the main threats of the AI revolution: the concentration of power. "There are two risks I worry about the most, which are a bit in tension," he said. "One is loss of control, that AI becomes too powerful and we cannot ensure control. The other is that power becomes too centralized, that there is one company, one model, or one person with too much power."
According to him, both scenarios are equally problematic: "In both cases, it is a very anti-human situation. The right approach is to say: we want people with deep control over the future. We want people empowered."
The fear of centralization is particularly interesting in light of the status of OpenAI itself, which in less than four years has turned from one of the research companies in the field into one of the most influential companies in the technology world. Altman said that at the beginning of its journey, in 2016, the founders of OpenAI also did not have a clear plan. According to him, the company's people gathered, bought a whiteboard, and tried to understand what exactly they should do. OpenAI, he said, actually spent about four and a half years before it reached a significant breakthrough – a path almost opposite to the accepted recipe in Silicon Valley, where companies are required to launch quickly, get customers, and change direction according to market feedback.
Even after the launch of ChatGPT, according to him, it was not clear to everyone in the company that this was necessarily the right direction. Some of the OpenAI people thought that the growth was unstable, and the company examined several other directions. Altman said that investor Peter Thiel was among those who encouraged him to focus on the chatbot, which he compared to a new interface similar to Google's search box.
And yet, even if AI continues to improve at a faster pace than the economy, Altman does not believe that humans will become redundant. When asked what will remain for humans in a world where AI is capable of performing almost every task better than them, he replied that the need for human connection will not disappear. From his perspective, even if artificial intelligence can create better content, conduct a better conversation, or perform work more efficiently, people will still want interaction with other people. In a world where more and more activities will become digital and automatic, precisely the human, physical, and authentic things may become rare – and therefore also more valuable.
Altman's central message, then, is that the problem of AI is not necessarily a lack of technological capability. On the contrary: it is possible that the technology is already advancing faster than society is capable of digesting. For OpenAI, which recorded a slowdown in the rate of revenue growth alongside growing losses, this is a convenient explanation, but also a significant admission: even those who stand at the forefront of the artificial intelligence revolution found it difficult to estimate how slowly humans and the economy would change their ways.
