Unlike specialized AI from programs like ChatGPT, this one is characterized by more capabilities and promises to “surpass” human intelligence.
Recent changes in OpenAI sparked a scandal among specialists Silicon Valley and those who are interested in technology .
In less than a week, the company of artificial intelligence (IA) has faced several changes within its board of directors, ranging from the dismissal of Sam Altman last Friday until the announcement of his reinstatement for signing this Wednesday.
Even if, beyond the official declarations, the reasons which led to this first decision remain unclear, Anonymous sources close to the company shed new light this Thursday .
According to what they told the news agency Reuters Speaking on condition of anonymity, Altman’s surprise departure as CEO of the company came after a series of advances in a project called Q* (pronounced Q-Star), which seeks to achieve so-called artificial general intelligence (AGI) .
That “could threaten humanity,” they suggested .
But, How is this AI different from others that we have already seen in programs like, for example, ChatGPT ?
What is artificial general intelligence (AGI)
There are different definitions that describe what AGI is or general artificial intelligence (AGI), as it is called in English a concept that existed long before the introduction of the recent OpenAI drama.
AI specialist Nisha Talagala, who is also CEO of AIClub, a company dedicated to AI literacy: wrote a column for Forbes in which he gave his own determination .
“I would define AGI as an intelligence that is not specialized in any specific task, as has historically been the case with AI . Today, most AI focuses on one problem and is extremely effective at solving that problem, often better than humans. For example, more than 20 years ago, an AI beat human experts at chess, but it couldn’t read a book, plan its day, or do anything else humans can do. .
Although there are different artificial intelligence programs that allow you to perform specific tasks, Talagala pointed out that The fundamental difference with the AGI is that the latter “could do all these things” .
Also from the aforementioned media ChatGPT, OpenAI’s best-known product, was asked how it would define this concept. .
The program responded that They are “highly autonomous systems that have the ability to outperform humans at almost any economically valuable job.” .
“AGI is often contrasted with narrow or specialized AI, which is designed to perform specific tasks or solve specific problems, but which lacks the broad cognitive capabilities associated with human intelligence. The key feature of AGI is its ability to generalize and adapt to a wide range of tasks and domains. “, added the program, according to Forbes.
in a conversation TED made last month, OpenAI Chief Scientist Ilya Sutskever shared his own views .
He said that The premise of AGI is that it would have the potential to be more intelligent than humans in all areas. and what’s more, he would have all the human knowledge to back it up.
Likewise, he stated that This would have the ability to self-learn, which would allow it to create other, more powerful AGIs. .
One of the concerns raised by some experts, like Talagala, is that This technology promises to revolutionize the job market much more intensely than previous AI .
For example, said that if an AI could read an x-ray and help a doctor make a diagnosis, an AGI would allow the system to read the x-ray, review the patient’s history, make a recommendation and explain to the person what is he suffering from?
Talagala suggested that If the above were to come to fruition, it could potentially replace a doctor in several aspects. human.
As expected, Contrary to this thesis, some specialists believe that tasks such as those of health personnel, to name just one, can never be replaced. .
The same Bill Gates co-founder of Microsoft, said in the middle of this year on his personal blog, Notes on doors .
Even like this, Analyzes on the future of AGI remain varied .
Winner of the Turing Award in 2018, Geoffrey Hinton – also known as “the godfather” of AI – said in an interview with CBS News who estimates that the AIG is less than 20 years away from being realized, while his fellow winner of the same prize, Yoshua Bengio, said The Next Web this still cannot be predicted how much is left for this.
However, There seems to be a general consensus that specialists should not be the only ones concerned about accelerating progress. of artificial intelligence.
Why should we care about this technology?
For Talagala »This week’s drama over OpenAI demonstrates how little we know about the development of a technology so fundamental to the future of humanity. and how unstructured our global conversation on the subject is.
In this sense, he declared that Questions need to be asked such as when will it be known whether the AGI has been achieved, what steps will be taken to regulate it and how it will be used by different factions. of the society.
In addition to what has already been mentioned, he stated that He doesn’t think there’s any need to worry much about the possibility that these systems will decide on their own to wipe out humanity. as seen in such popular films as Terminator.
“History has shown us that the harm caused by technology is often due to intentional or accidental misuse by humans. It is possible that artificial general intelligence will achieve some form of consciousness independent of humans, but it seems far more likely than human-directed weapons, disinformation, job cuts, environmental disruption, etc. threaten our well-being. of this”.
This is why he believes that It is crucial that citizens are aware of what is happening with new technologies so that people can exercise their rights and strengthen their opinions.
“Technology is transformative. What is not clear is who will decide how it will be transformed. “.