Solving Solved Problems or AI as a Tax Dodge

The Verge has an article laying out the debate around the use of imitative AI in the law. There is some discussion around the use of imitative AI to parse technical documents in arbitration cases, a discussion of a judge who uses it for research, and a discussion of the many, many example of imitative AI introducing bullshit into legal filings and decisions. All of that is interesting, and the story is well reported, well worth your time. And none of that interests me, not really. What this story crystalizes is just how AI is meant to insulate rich people from taxes.

I want to be clear that many of the people in this story have absolutely the best intentions. There really are too few judges in the system. Cases really do take too long to get through. Money really is a deciding factor in far too many cases. Arbitration, especially technical arbitration, really does require a lot of effort to parse. The people trying to solve these problems are trying to solve real problems. The issue, then, is not their intentions, but their solutions. All of the problems that they are trying to solve has a solution that is more reliable and has many fewer externalities: hire more people.

All of the problems mentioned above could be solved with a combination of more labor and better processes. Hire more judges. Hire more experts for arbitration. Hire more people to process briefs, etc. This reasoning applies to many public service areas where imitative AI is pushed as a solution. Having a hard time giving kids the education they need? Pay teachers better and hire more of them. You could do all that for likely much less than the true cost of these systems (one thing not covered in the article is that imitative AI is not making money and cannot make money at the prices it charges. The cost savings you may see today both do not take into account externalities like climate change and are illusionary) and you wouldn’t burn the planet in the quest for time savings. Win, as they say, win.

But!

But there isn’t enough money to hire all the people we need! Well, yes, yes there is. If we make the rich people pay a fair share of taxes, there is more than enough. We have a tax issue, not a money issue. We decided, as a country, that we wanted the richest among us to pay as little as possible — the richest actually pay lower taxes than the average of all Americans. We have, in the past, taxed the most well off, the people who benefit the most from a strong, stable society, enough to create that society. We could choose to do so again. In fact, we are doing so already, in fits and starts. California is going to vote on a one time wealth tax, and Massachusetts’ millionaire tax has brought in more money than expected while the number of millionaires in the state has increased. People openly discuss universal basic income and job guarantees. The tide has been shifting toward a more responsible society for some years, and imitative AI is one means of trying to hold back that tide.

The people who push imitative AI are generally the richest of the rich — founders, tech leaders and CEO, finance titans. They certainly, as a class, do not want to go back to the days of a strong society if the cost is them paying more taxes. Those yachts and politicians aren’t going to buy themselves. So imitative AI is pushed — no need to build the human capital required for a strong society. No, imitative AI can do it all for you, if you just ignore the climate damage, the theft needed to train it, the child sex abuse material it produces, the supercharged disinformation, employment disruption, and the fact that they cannot charge prices that produce profits. Imitative AI is not entirely about hoarding all the money, but that greed certainly contributes. Notice how UBI is never mentioned by these folks anymore? Sam Altman’s sentences used to consist of a noun, a verb, and a UBI but now he doesn’t even use words containing those letters.

We are a society. Everything great that we have achieved — in art, politics, medicine, science — is because we invested in the human capital required to build a strong society. We had a deal: people who worked hard and got lucky in their field could be wildly rewarded, give be handed fsck you money if they contributed some of that money back to endure the society could produce the next generation to follow them and ensure the people who built the infrastructure they stood upon had decent lives. It was a new deal in history, one that Reagan and his ilk began to dismantle in the 80s. Today, we have people trying to cash out on a money-losing tool in the hope that they can avoid having to accept that deal once more. The rest of us, the ones that make a strong society and their success possible, don’t have to accept that betrayal.

When someone starts claiming that imitative AI can solve the problem of a fraying society, ask them why we don’t just hire the amount of people needed to fix the problem. When they say we cannot afford yo do so remind them that we can. We know where the money is; we just need go get it.