Real intelligent behaviors necessarily lead to more trust.
Trust is evaporating everywhere in this digital era.
Forgot to bring your phone to work? You’d better go back home and pick it up, or you won’t be able to authenticate your 2FAs. Moving to another country? Better not cancel your phone number, or you’ll lose access to accounts requiring text verification. Are we living in a more secure and trustworthy society compared to decades ago? Not really. The reality is that you must bow to the digital overlord to prove that you are indeed yourself.
Are humans any better? I spent two hours this morning trying to prove my identity to customer service because I forgot to save the “passkey” on my phone and needed to recover access to my account. The result? No luck. I complied with every command: shaking my head, saying three random numbers in a video recording, holding my IDs and a piece of paper with today’s date and location, and even taking off my glasses. They still rejected my request, finally saying the ID must be issued by a US state rather than a foreign government. My guess is that some AI algorithm on the back end didn’t recognize my foreign ID, and the human service decided not to bypass it.
Ironically, as a computer science professor and researcher, cryptography and AI have just denied my humanness. Seek legal assistance? Probably not worth the money for a single account. The fundamental issue is that we now live in a society where technological and human stupidity make things worse for everyone.
Here’s the big philosophical divide between cryptography and intelligence: Cryptography tries to hash and distinguish tiny differences between two copies, whereas intelligence tries to make sense of small nuances and merge them into a single concept. Take two photos of the same ID. A hashing algorithm will say the two pictures are different because all the pixels have changed their values. For security reasons, no value is allowed to change. But to our eyes, we know they are the same. And not only that, if you show me another picture of the same person, I could tell they are the same person right away (approximately).
When my family and I first immigrated to the Western world, one of my first impressions was that there was more trust here than where I came from. People didn’t seem so nervous about verifying everything back then. There were no gates in subway stations; people didn’t lock their backyards; and border agents even accepted an oath of citizenship to enter a country. There could be a hefty penalty for fraud, but we trusted strangers. Now all of that trust is gone.
In a trustless society, everyone must prepare to waste more resources on verification — taking out your cell phone and entering one-time passwords every day, taking off your shoes and bringing out laptops at the airport, and being prepared to present all forms of IDs and birth certificates to prove you are yourself. The crypto community is proud of their construct of a “trustless society,” but the proof-of-work framework in cryptocurrency and blockchains consumes hundreds of billions of kilowatt-hours of electricity per year, turning it into heat, and duplicating tens of thousands of copies of the entire blockchain in every single node, just to mathematically “prove” every single transaction.
Real intelligent behaviors necessarily lead to more trust. Real intelligence values the time and resources it spends and recognizes that trust can bring more efficiency for all. Real intelligence measures the value lost as a society when trust is lost and weighs long-term benefits over short-term profits.
In the past, human stupidity has led to constant upgrades of security measures. Some people commit crimes for their short-term selfish profits, diminishing other human beings in their value equation. But the insanity happening now is a rivalry between human and technological stupidity. While we crank up more technology to prevent crimes, the holes in the technology itself cause more harm and waste more resources. I deeply worry that, in the short term, rapid deployment of dumb AIs will create more harm for people who do not fall into the norm — minorities, people they haven’t been trained on, or people who simply hold foreign passports. AIs that lack real intelligence will ignore rare cases, cannot adapt flexibly or reason thoroughly, and are unable to handle uncertainty. The scary thing is that human agents will follow whatever the AI says and deny the 1% edge cases to escape their responsibility — a world where AI becomes our de facto ruler.
At the end of the day, you can’t really prove that your biological body is truly yourself, can you? :wink