Google's CEO "doesn't sleep at night" worrying about Artificial Intelligence
Its CEO Google, Σούνταρ Pitσάι, δήλωσε πως οι ανησυχίες του για την Τεχνητή Νοημοσύνη «τον κρατούν ξύπνιο τη νύχτα», προσθέτοντας πως η τεχνολογία αιχμής θα μπορούσε να αποδειχθεί «πολύ επιβλαβής» αν εφαρμοστεί λανθασμένα.
Pichai thinks so Artificial Intelligence will require a global regulatory framework, similar to that for nuclear arms control. The head of Google acknowledges that competition in the technology industry could set aside the concern for security issues.
In an interview on CBS's 60 Minutes, Pichai said governments will need to develop global regulatory frameworks alongside the development of AI systems.
About two weeks ago, the question of the potential dangers of A.I. came back into the public sphere when more than 1.000 researchers and tech leaders, including Elon Musk and Apple co-founder Steve Woniak, called for a six-month "pause" in the development of powerful A.I. technology. In a public letter , they said this technology could be accompanied by "serious risks to society and humanity".
Google's parent company, Alphabet, owns UK-based artificial intelligence firm DeepMind. Google has already launched an AI chatbot, Bard, in an attempt to 'answer' the popular ChatGPT and OpenAI's multi-billion deal with Microsoft.
Pichai points out that artificial intelligence could cause problems, because of the potential for misinformation. «Με την τεχνητή νοημοσύνη θα είναι εφικτό να δημιουργηθεί εύκολα ένα βίντεο, όπου θα μπορούσε να φαίνεται ο Scott (σ.σ Σκοτ Πέλεϊ, ο reporter του CBS που του πήρε τη συνέντευξη) να λέει κάτι, ή εγώ να λέω κάτι, που ποτέ δεν είπαμε. Και θα μπορούσε να είναι αληθοφανές. Σε κοινωνική κλίμακα μπορεί να προκαλέσει μεγάλη ζημιά».
The head of Google has argued that the version of its artificial intelligence technology that is now available to the public, through the Bard chatbot, is safe. He added that Google is acting responsibly, not releasing advanced versions of Bard to the public for testing.
NYT: Safety Sacrificed for Speed
Just 24 hours ago, the New York Times revealed that Google launched Bard after years of internal wrangling over whether the benefits of A.I. outweighed the risks. In 2020, the tech giant also announced Meena, a similar chatbot, but that system was ultimately deemed too risky to launch, three people with knowledge of the matter said.
These concerns were revealed earlier by the Wall Street Journal.
"Google was aware"
According to the NYT, Google blocked two of its top AI and ethics researchers, Tinit Gebru and Margaret Mitchell, from publishing a paper warning that the large language models used in new A.I. systems , they could unleash abusive language.
In January 2022, Google tried to prevent another researcher, Dr. El Mahdi, from publishing a critical study, according to the NYT.
"We don't even fully understand our own AI system"
Now Pichai admits that Google didn't fully understand how its own AI system responded in certain situations.
"There is an aspect of this phenomenon that all of us in the field call the 'black box.' We don't fully understand it. And we can't say exactly why he said something or why he made a mistake."
Asked by the CBS reporter why Google released Bard while not fully understanding how it works, Pichai replied, “Let me put it this way. I don't think we even fully understand how the human mind works."
Pichai acknowledges that society doesn't seem ready for rapid advances in artificial intelligence. "There seems to be a mismatch" between the rate at which society thinks and adapts to change compared to the rate at which artificial intelligence is evolving. However, he added that "at least people are starting to realize its potential dangers more quickly."
“Compared to any other technology, I've seen more people worry about it earlier in its life cycle. So I feel optimistic."
Pichai stated that the economic impact of Artificial Intelligence will be significant and will affect everything. "Every product in every company."
Αναφερόμενος, για παράδειγμα, στον κλάδο της ιατρικής, ο Πιτσάι είπε ότι σε πέντε έως 10 χρόνια από today, ένας ακτινολόγος θα μπορούσε να συνεργάζεται με έναν βοηθό Τεχνητής Νοημοσύνης, για την ιεράρχηση των περιστατικών. Ο ίδιος εκτιμά πως μεταξύ των επαγγελμάτων που θα επηρεαστούν είναι οι συγγραφείς, οι λογιστές, οι αρχιτέκτονες και οι προγραμματιστές.