Two weeks ago, in a post entitled “When will Academia stop focusing on what the industry already does and start focusing on what the industry does not want or cannot do?” i wrote that in the medium term companies will inevitably end up dominating technological research on this planet, or at least that research from which they can extract high profits.
Interestingly, in the last issue of The Economist, in which the core theme is dedicated to the latest developments in the field of artificial intelligence, one can read that not even one of the richest universities on the planet (Stanford University) can build Nvidia supercomputers and that this area ends up dominated by the interests of large private companies, the only ones with the financial capacity to build supercomputers whose training costs 1 billion dollars. The “deadly” sentence in the article is: “Academic institutions can no longer keep up”
In this context, it is important to remember that just 2 days ago Google (whose market capitalization is over 1 trillion dollars and is one of the few companies on our planet where some top Engineers can actually have an annual salary of over 1 million dollars, who by the way deserve them much more than soccer monkeys) suspended one of its Engineers, who made worrying statements about Google´s artificial intelligence. But the scenario described in the aforementioned article is not such as to dissipate those risks. In fact, the article admits that this scenario can become a reality and that Google is the company that can achieve it. I am not referring to the content of the article by another Engineer from the same company, but to statements made by one of the leaders of Eleuther.
Pacheco-Torgal 