The promises and risks of AI dazzle and worry in Davos | Technology

The astonishing evolution of generative artificial intelligence is a transformative force of impossible-to-measure importance and risks, but it shows signs of becoming one of the most revolutionary leaps made by humanity. Despite the terrible wars that require attention, political, economic and opinion leaders paid great attention to the issue at the Davos Forum, in its economic dimension (with the increase in productivity and the impact on the market of work), regulatory (the opportunity to emphasize protection or innovation), geopolitical (as one of the decisive elements in defining the balance of power) and scientific (with the opening of previously inconceivable borders ). A terrain full of uncertainties and dilemmas, in which the only certainty is a level of transcendental importance, which has been clearly expressed in speeches, panels, hallways and private meetings.

The uncertainty about the direction this technological revolution will take is great, and Sam Altman himself, CEO of OpenAI, the company that created ChatGPT, assumed this during a public session. “This technology is very powerful and we cannot say with certainty what will happen. This happens with all major technological revolutions, but with this one in particular, it’s easy to imagine the enormous effects it will have on the world and that it could go very wrong. We are going in a technological direction that we think is safe, but I understand the concerns,” said Altman, an industry authority. Altman warned that “stress will increase as we get closer to AGI, artificial general intelligence, capable not only of performing specific functions such as language models, but also of learning any intellectual task.

Faced with this scenario, the regulatory debate is one of the most delicate. In the same Altman panel, Jeremy Hunt, British Chancellor of the Exchequer, was in favor of “light” regulation, thus positioning himself at the pole of those who consider that regulatory frameworks which stifle innovation in a sector endowed of enormous potential for improving productivity and enabling decisive possibilities also on the scale of geopolitical competition.

AI is undoubtedly a central element in defining the forces of the future. In the Davos proceedings, Europeans’ concern about being left behind in this revolution emerged. The EU is a pioneer in regulation, but it is not at the forefront in terms of leading companies in the sector.

The balance of power that will define AI is measured in terms of economic benefits for businesses, but also in terms of its ability to ensure the least disruptive change possible to the labor market. An IMF report released on the eve of the forum’s start indicated that up to 60% of jobs in advanced economies could be affected by the emergence of AI, with half of them negatively affected.

Many jobs will disappear. Others will appear, but not necessarily at the same time, and certainly not necessarily to the same people and places. Mitigating the pernicious effects of the revolution will be a measure of the cohesion and stability of future societies.

The promise of scientific advancements that underpin the AI ​​revolution is also enormous. Alber Bourla, CEO of Pfizer, made this clear in Davos. “This revolution is transforming the organic sector,” he explained. “We use AI very intensively. Thanks to this we achieve better and faster results. Previously, the drug discovery process typically took four years. We would synthesize millions of molecules. Today, with AI, we have moved into drug design. We make around 600 molecules, chosen with enormous computing power, and which are most likely to work. The process of years became months. “It’s something that saves lives,” he said.

On the worry side, without getting to the apocalyptic visions of computer systems of superhuman intelligence taking control, there are much closer and much more real risks. One of them is the potential of this technology to increase disinformation activities, for example in electoral processes.

Disinformation and the threat it poses to democracies have emerged as one of the two biggest risks facing the world, according to a report released by the World Economic Forum on the eve of the Davos meeting. Generative AI can harm in two ways: a quantitative one, allowing the mass creation of content without a human being having to do it; another qualitative one, with counterfeits of such extraordinary quality that the capacity for persuasion is total. In forum discussions, concerns arose, for example, about the video variant of this risk.

The fears are of sufficient magnitude that, as reported in the newspaper Financial Times A week agoAmerican experts – notably from Open AI – and China held two secret meetings to address the risks of disinformation and the threat to social cohesion.

The news takes on an extraordinarily worrying aspect if we consider that the United States and China are engaged in fierce competition in technology, and in particular in AI, which is a source of serious friction between them. Washington is carrying out maneuvers aimed at restricting exports of advanced electronic chips which are necessary to move forward on this path and for which China does not have independent production capacity. Washington claims this is justified to prevent Beijing from using Western technology to power military and security programs with highly dubious objectives.

In Davos, Chinese Prime Minister Li Qiang came down hard on these maneuvers, trying to establish a link between them and a general attempt by the United States to maintain a privileged position and complicate the access of emerging countries to technologies keys. US National Security Advisor Jake Sullivan responded that it was not a general blockade, but only a specific measure.

That in the midst of the tension between the two powers, there was a contact like that described in the FT He is very eloquent.

Another problem that emerged at Davos is the staggering energy consumption required by new computer systems. An estimate published last year calculated that by 2027, AI servers could consume as much energy per year as countries like Argentina or the Netherlands do in a year. This overloads demand and, in countries that do not have good levels of green energy production, leads to more CO2 emissions.

The AI ​​revolution will touch almost every aspect of life, likely including, as noted In a recent interview with this newspaper, historian Niall Fergusonour cognitive capacity, which can be affected by the future, constantly turning to a solution machine rather than our own thinking.

You can follow EL PAÍS Technology In Facebook And X or sign up here to receive our weekly newsletter.

_