Portuguese scientist Tiago Ramalho, a former Google DeepMind Technologies engineer, said in an interview with Lusa that “within ten years” the world population will no longer “imagine what life is like without artificial intelligence.”
The CEO and co-founder of Recursive, a Japan-based startup, believes that because “the field is moving so fast,” within five years, artificial intelligence will be “embedded in most products and services.”
Within a decade, artificial intelligence will become “a ubiquitous technology,” Ramalho said, starting with natural language processing, including ChatGPT and other models capable of generating complex text.
The Portuguese defended that, in the last two years, technology has reached a point where “it even works quite well” and evolution has become “an engineering problem”.
In other words, Ramalho explained, the challenge will be how to increase the number and quality of the databases, reduce the costs of obtaining a response and adapt the models “perhaps even for mobile phones.”
According to a study published in March by Microsoft researchers, the popularization of artificial intelligence will affect professions mainly in the areas of telemarketing, accounting, translation, teaching and programming.
“When there is a technological transition, there are always people who will be harmed and others who will benefit,” admitted Tiago Ramalho, who believes that “the end result always tends to be positive.”
Some scientists believe that a sentient artificial intelligence could be another danger. One of them, Google engineer Blake Lemoine, said in June 2022 that the tech giant’s LaMDA language model was “alive.”
Google rejected Lemoine’s accusations and also Tiago Ramalho believes that LaMDA is simply “responding by following the context it is receiving” and “doing a bit of staging.”
“If we ask him ‘are you suffering locked up in this data frame’, he will answer like in a story by Isaac Asimov (1920-1992),” said the Portuguese, referring to the science fiction writer.
“There is still a long way to go” until the creation of an artificial intelligence with “an internal dialogue, with preferences and values that do not depend on interactions with human beings,” Ramalho defended.
The scientist believes that “the most realistic thing is for humans to co-evolve with technology”, building a society with the benefits of “a symbiosis between intelligent silicon agents” and humans.
Ramalho gives Google DeepMind’s AlphaFold technology as an example, which “allows us to understand the structure of a protein from the genetic code.” “This is a problem that a human being simply cannot solve,” he said.
Recursive, currently with around 40 employees, is also using artificial intelligence to help Japanese companies create sustainable business models, the Portuguese explained.
The startup is, for example, optimizing home deliveries for a group and forecasting the irrigation needs of commercial forests managed in Indonesia by the Japanese conglomerate Sumitomo.
“We always try to find that synergy between what is good for the environment and also for the company, because sometimes it is difficult to sell the idea of sustainability,” Ramalho explained.
Source: TSF
