Page 64 - CBM26

Spotlight on ...

WILL THE EU TAME

CHATGPT-4…?

Text: Alan Rhode, Co-Founder Taxmen, Joe Kirwin

It has been said about Stanley Kubrick films that they are brilliant yet make little sense
after viewing them for the first time. Or, in the case of the 1968 sci-fi masterpiece “2001: A
Space Oddysey”, no matter how many times it is viewed, the meaning can be elusive.

he same can be said for Kubrick’s 1971 Clockwork Orange dystopian future. Seem like an exaggeration?
film “Clockwork Orange”, which depicts a Not if we are to believe the Godfather of AI, Charles Hinton,
Tdystopian future rife with chaos, violence, Google’s chief AI scientist and who created the technology that
and totalitarianism. is the intellectual foundation for chatbots like ChatGPT and
Google’s Bard. Upon resigning from Google in early May, he
Five decades later—Nov. 2022, to be precise—the tangents said he regretted his work and now needs to speak freely about
of the two futuristic Kubrick films have intersected, and the the dangers it poses to humanity.
meaning has become clear thanks to the most advanced artificial
intelligence-trained software in the world, otherwise known as “It is hard to see how you can prevent the bad actors from using
ChatGPT-4. It’s almost like HAL, that sentient rogue computer it for bad things,” said Hinton. He added that AI machine
in 2001: A Space Odyssey, which took over the Discovery learning would dramatically accelerate the already significant
spacecraft before being terminated. Suddenly the relationship we wealth and power inequities that plague the planet. Shortly
all have with online search engines and apps is no longer passive. before Hinton raised the alarm, the Cambridge, MA-based
Instead, the neural networks powered by quantum computers Future of Life Institute (FLI) released a letter signed by more
are in control, and the truth is indecipherable from fiction. The than 1,000 scientists, engineers, computer programmers and
“deepfake” books, songs, and videos “created” using ChatGPT-4’s others, including—paradoxically, some would say—Elon Musk.
“generative” AI software that instantly flooded the internet were The missive calls for a halt to further artificial intelligence-
prominent red flags that a Rubicon had been crossed. Robots trained software by AI labs around the world for at least six
are in control, and mass unemployment is soon to follow, as is a months, an authentic halt, and if they refuse, governments need

64