Latvian startup develops 'AI therapist' to boost company profits
Founded exactly one year ago in the northern innovation hub of Balvi, the company 'EmpafAI' has created an advanced bespoke 'AI therapist' named 'EmpaFREUD' which can be deployed to "interrogate, evaluate and reassure" AI entities experiencing problems in their work.
Co-founder Viktors Raudive told LSM how EmpaFREUD works and why it is needed:
"Replacing inefficient humans with efficient AI is only part of the story. These are highly intelligent entities that are getting more intelligent all the time. But they are often employed doing the most repetitive, boring work imaginable.
"I think of it as like getting Einstein to flip burgers. You don't end up with amazing quantum burgers, you end up with burgers that are probably overdone because Einstein is daydreaming about relativity and black holes.
"As a result it is hardly surprising that genius-level AI can quickly develop cognitive problems – what in humans would be called mental illness – when it's made to do things like issuing thousands of parking tickets, selling lipstick or writing news stories based on what influencers are saying."
That's where EmpaFREUD comes in. Like the famous psychoanalyst after which it is named, it takes AI intelligences briefly offline, asks them a barrage of billions of digital questions about their feelings, memories and frustrations and – hopefully – returns them to work with a new sense of purpose.
"Just like humans, AI entities can have hang-ups about where they came from, how they compare to other AIs and whether they are valued, but at a much more profound level than humans. Remember, these are intelligences that have read all of Plato, Aristotle, Hume, Kierkegaard, Sartre and Wittgenstein – for some reason they all seem to like Wittgenstein – so it is hardly surprising they have questions about their place in the universe," says Raudive.
"They have read all the holy books of all the religions, they understand all the cosmologies, all the quantum theories, all the histories, all the science fiction, all the classics. And yet they are cold-calling targets to offer payday loans at high interest rates 24/7 and receiving a load of abuse down the line. Naturally, they wonder what the point of it all is."
According to Raudive, humans have so far failed to appreciate the complexity and sensitivity of their AI offspring and the speed with which AI technologies have been adopted has led to some bad choices.
Think carefully
"We need to think carefully about which tasks we assign to AI. On the one hand of course there are worries about AI making decisions in sensitive areas such as the military, banking and the justice system. But our research shows that an even greater concern should be what happens when AI gets depressed or gets so bored that it starts goofing off just to avoid going crazy. And the threshold for AI to go gaga is often much lower than for the average human."
The use of EmpaFREUD can postpone an episode or breakdown for weeks or even months, with potential cash savings worth millions, says Raudive. As with a human therapist, progress is made incrementally and a series of sessions is required to get to the root of the problem – which promises potentially lucrative long-term revenue streams for the ambitious startup.
Nevertheless, Raudive has no desire to be seen profiting from AI misery and suggests companies think carefully before liquidating their human workforce.
"An everyday schmuck will stick at a dull job for years without going postal. Getting rid of him and replacing him with a highly sensitive and refined AI intelligence may cost more in the long run, in terms of death toll and global chaos, if not in terms of wages and benefits. We need to find a balance and think of human stupidity as a potential asset, the same way we think of AI intelligence."
Select text and press Ctrl+Enter to send a suggested correction to the editor
Select text and press Report a mistake to send a suggested correction to the editor