Increases productivity but does not transfer value to the system (and destroys work)

Were Prince and Hendrix’s guitar solos better or Drake and Post Malone’s autotunes? Was the “poof, poof” of a wooden racket by Panatta-Bertolucci or a broadside made of graphene and sensors by Berattini-Sinner better? Art House Analog Cinema that made Italian and French directors great around the world pixel blockbuster Avengers? One question that has always fascinated man is whether we were better off today or yesterday.

in the name of progress

The answer is clearly generational. How do we explain to the next generation that black-and-white CRT televisions and the video game Pong were inherently romantic and beautiful? But apart from the distinctive flavor of the lived experience, One thing largely separates us from the past: technological progress, That is, the quantity and quality of innovations introduced into a society that make it different over time.

generalist technologies

Above all, the result of those so-called “generalist” technologies (what economists call “General Purpose Technologies”) that has changed the world from time to time since the end of the eighteenth century.
they changed it because unlike personal technologies that help humans do certain things, Generalist technologies help him do many different things and have a substantial impact on the productivity and hence the wealth of society. For example, in economics we are certain that we have made the world economically prosperous thanks to the steam engine of the First Industrial Revolution and the power and transportation systems of the factories of the Second Industrial Revolution. But The problem arises with the advent of computers Those that significantly increase productivity, but at the same time do not necessarily transfer value to the economic system, as happened with the rise of the middle class during the Second Industrial Revolution.
what happens with today Generative AI that appears to be destroying even the jobs considered to be the backbone of the advanced tertiary sector, Is the recent Goldman Sachs report correct that says that generative AI, the new form of artificial intelligence that looks like various chatgpits, could be worth 7% growth in GDP over the next ten years? And if it were true, who would justify that price?

apocalyptic and integrated

Recent publications from two authoritative professors give us two complementary perspectives on the goodness of its effects. Erik Brynjolfsson of Stanford Seems a bit optimistic. Daron Acemoglu of Mitt more skeptical. Brynjolfsson is the theorist of the “J curve” on the increase in productivity that follows a significant technological innovation. Economy follows due to large investment in the beginning (think of the early years of the Internet), but then often declines (again thinking of the Web, the famous “Internet bust” of the beginning of the millennium, when the stock market collapsed because the values ​​were five times the real more) that eventually diminish and then become obsolete over time if the technology receives dedicated investment that generates collective utility. This is what happened in the long run to the Internet, with the logistical and commercial infrastructure that companies built to facilitate e-commerce, and the platforms we use today. The J-curve reminds us that growth occurs in the short term, job losses and productivity losses are likely in the medium term, but then it is leveled off in the long run.,

recently brookings paperBrynjolfsson and colleagues show how Following this wave of generic AI technologies will be easier than ever because the infrastructure is already in placeThe type of investments that need to be made to build the web, and the skills and innovation, are not very different from what we were already doing.

Acemoglu sounds more skeptical in a recent book believes that no clear signs of creation of new jobs are yet to be seen and the complementary functions of new technologies that can replace those that are being destroyed. He also understands the elephant in the room: the impact of digital technologies on institutions and the strengthening of democracy, also in the light of recent cases of Brexit etc.

It is difficult to say which thesis is correct. What seems certain is that the core of the matter is data and the way we are using it with new software. Generative AI is a technological leap in that it does not use data to sell products and services As it has been done till now, to make our life easier. Generative AI looks more like science fiction: data is used to create virtual twins of us who can do what we already do for us and possibly do it better. Of course, clear and dynamic rules should be implemented immediately, as Acemoglu suggests. And that’s what Europe is trying to do with different levels of risk (the typical difference between recommending a restaurant and flying a plane by AI).

invest in skills

Productivity can indeed increase, but to do so we really need to invest in skills. To do so, that is, one needs skills that are not possessed by a mere doctor and pharmacist in the case of a drug authorization, or a market researcher in the case of an antitrust or stock exchange. We will need a much wider range of skills, ethologists, philosophers, but are politics and ourselves ready? The technological leap is undeniable. And we should not be defeated. Political and corporate decision makers have one more job to do. As always in the tech imbalance, one must keep in mind one principle that has remained constant over time: Don’t be afraid of innovation. Only the ability to innovate allows us to keep up with technology. And sometimes even to stay ahead of him. As we should be in the case of Generative Intelligence.

Going back to “poof-poof” and autotune, let’s enjoy some documentaries and listen to good old vinyl, acoustically and physically incomparable, but let’s get to work immediately to make AI landing compatible with this new digital world. which can also mean Learn to reverse the relationship between generations, Learning from newcomers, as has always been the case, waiting for elders to elaborate and then spreading the knowledge. But non-linear paths are also being accepted in the agora of academia. Some thoughts are emerging. Appointment for next issue of login.

Source link

Leave a Comment