For those hoping for the rapid advancement of artificial intelligence, Sam Altman’s victory in keeping his position as CEO of OpenAI was also a victory.
A debate surrounding AI development centers on OpenAI and Altman: how quickly should humanity strive to create “artificial general intelligence”—that is, intelligence that is fully humanlike or possibly even superhuman?
There was, as it were, a new pope of AI this past week at OpenAI, which almost caused a schism. Emmett Shear would have been this new leader of OpenAI, the field’s most powerful organization. Like Altman, he was an early member of the venture firm Y Combinator and co-founded Twitch. The board appointed Shear as interim CEO after dismissing Altman last week.
With Shear leading OpenAI instead of an aggressive leader like Altman, the once-sizzling startup would almost certainly have slowed its rapid pace of development. In September, Shear wrote on X about a proposed pause in AI development, saying, “If we’re at a speed of 10 right now, a pause is reducing to 0. I think we should aim for a 1-2 instead.”
It is challenging to pinpoint an individual’s exact position on the wide range of AI safety concerns. On these issues, Altman disagreed with Ilya Sutskever, co-founder of OpenAI, and chief scientist.
Had Altman moved to Microsoft as CEO Satya Nadella had stated he would after being fired, he would have effectively joined big tech in commercializing artificial intelligence, thereby transforming into a sort of antipope.
All others would have to choose which of these groups—or at least the camps they would represent—to be loyal to, whether they were businesses attempting to use OpenAI’s technology or individuals who have already grown dependent on it.
Instead, in a time when tech leaders and the products they create are idolized in the secular world, Altman can resume leading from San Francisco, which is as close to Vatican City as we’re going to get in this era.
After a decade of covering this topic, almost all the experts interviewed have told that we are nowhere near achieving artificial general intelligence, in large part because we have such a poor understanding of how to build any intelligence, artificial or natural. This makes the schism a theological matter. According to Altman, he is unwilling to predict when artificial general intelligence will be achieved.
However, some in Silicon Valley believe that since artificial general intelligence is so close to becoming a reality, we ought to at least be thinking about the deep and wide-ranging implications of its origins.
This gives many people who are interested in this technology great anxiety, as they believe that an AI that is more intelligent than humans could go rogue and attempt to wipe out or subjugate humanity, as depicted in movies like The Matrix and Terminator.
This kind of behaviour is known as the “superalignment” problem, and it is exactly what a group at OpenAI led by Sutskever—the board member who fired Altman at the time—has been attempting to ensure future AIs don’t do.
Some who are pushing for the fastest possible development of AI claim that it is essential to help us resolve the most important problems facing humanity right now. This group includes people who call themselves “accelerationists” and make fun of people who are worried about AI safety by calling them “decels.”
Many of those accelerationists, some of whom are also extraordinarily successful capitalists, claim that the advancement of AI can benefit a great number of people worldwide who are in need. They argue that it is beneficial for AI to develop quickly.
So, should we applaud Altman’s comeback to becoming OpenAI’s leader or should we throw in our support on accelerationism’s victory?
That is impossible to answer because the very notion that we are getting close to artificial general intelligence is based on ideas that are outside the realm of current science and personal experience.
However, his return benefits OpenAI’s investors and highly compensated engineers, as well as those who wish to see the current generation of AI advance. With Sam Altman heading OpenAI once more, it looks like AI development will move much more towards “full steam ahead” than “wait and see.”
