There’s far too many referees, not enough players speaking up on the subject of AI.
To see congressional first drafts of legislation is a joke. They’re so unprepared and lack a fundamental understanding of AI, that we can not leave it in their hands. We are rightfully terrified of the inequality, oppression, and violence that will come from AI left in the hands of men, Tech Barrons. Yet, because regulators understand so little, intimidated or stupefied, they’re handing the keys right back to Tech Barrons to keep driving.
And all the while humans have enormous potential to be caring, generous, and compassionate.
So, what’s the point of AI anyways? Why should we want AI? What are we optimizing for?
In the first wave of AI, we saw models that optimize for productivity — speed, accuracy, and efficiency. Think robotics in operating rooms and language models to connect you to the right support in a chat bot.
In the second wave of AI, we’re refining for profitability. Think about models that use the Youtube video you watched or the book you recently purchased to target advertisements at the right time.
In the third wave of AI, there is an opportunity to refine models for quality. I build AI to improve the human condition that promotes free-will. My work at LOOP, uses AI to finally eradicate redlining from insurance. Let’s call it egalitarian-AI, one that presupposes access to opportunity as a right, not predetermined by gender, age, race, or starting economic circumstance.
These truths we hold to be unalienable, that all men are created equal. What if the math could also prove that, identify where inequality creeps in — whether that be where you live and your access to things like high quality education or environmentally non-hazardous conditions (e.g., lead in the water). And pinpoint ways to correct for this and create a more perfect union with policy, with investment, and with the entrepreneurial community.
Orienting models to improve quality is a little fuzzier because ‘quality’ it’s a matter of personal taste, and requires a tactful artist to produce. Its loose definition leaves it open to and vulnerable to exploitation. But there are people in the world (myself included) that are making extremely good, high quality models. People who can conceptual and create AI (the dangerous) have an opportunity to share this gift for the betterment of society, but we have to chose to create these models and share high quality work with the world (the generosity).
The third wave, of egalitarian, generous AI is possible if we orient our ambition to it.