The site uses cookies that you may not want. Continued use means acceptance. For more information see our privacy policy.

The Artificial Intelligence that is a Corporation.

What is climate change and carbon pollution but the gray goo problem?

Let’s start by asking what the purported dangers of general artificial intelligence are. The ones most listed include things like:

  • They won’t care about people and will therefore get lots of humans killed or kill them.
  • They won’t think about the consequences of their actions, instead blindly following whatever their goal is.
  • They will spread misinformation, and people will believe it. Those lies will cause harm.

Remind you of anyone?

Corporations (or any organization) are artificially intelligent constructions. While nominally there are humans at the controls, for a variety of reasons this is provably false and nothing but a happy fantasy.

You see, all those lawyers and C-suite suits are constrained by their directives and their personas (for which the corporations selected them in the first place), so they don’t really have all that free will stuff you’ve been taught since the Garden of Eden days.

If a CEO finds out that their company’s pollution is killing people, how many step up and shut it down? (I’d like to know, but couldn’t find a reliable source, so take it rhetorically if you wish.) More likely, they worry about the stock price, their child’s reputation, and they at best try to clean things up a bit, maybe pay for a few funerals.

The free will of any man is tempered by her fetters. She will only act as freely as she feels she is able. Only as freely as she is to think, and thinking is constrained by the limits of the information she has. She will not see through lies she prefers to be true if it’s more profitable to ignore them. We see this example repeated. Read about the meltdown at Chernobyl. Read statements from Republicans after they leave office and don’t have to spout schlock to get reelected. Over and over we see the pattern of the chains on thinking and on acting.

Now a lot of (rich and powerful) people share a curious worry about artificial intelligence. They call for rigor to ensure AI doesn’t discriminate against people (while various systems do just that in jails, in polluting of neighborhoods, in unequal education, and on).

They call for protections against robots used in war, while the man-made wars continue to litter the earth with indiscriminate weapons, like land mines and unexploded cluster bomblets, that will kill for generations after the war ends.

And as for misinformation, corporations lie and mislead with impunity as long as it’s not outright fraud (and if it is, they pay their fine and keep on existing). There’s even a whole sub-industry dedicated to misinformation, known as marketing.

Should we be worried about general artificial intelligence, and even lesser forms, being used to harm our society and harm our planet? Yes!

But, we should be alarmed (and many of us are) at all the existing ways the exact things are already being done. These threats are not novel to machine systems! They already happen. Most if not all have been going on for longer than the age of the United States of America.

The pollution isn’t worse because an AI does it. The discrimination isn’t worse because an AI does it. The deaths from indiscriminate warfare are not worse because an AI does it. The fact that an inhuman being perpetrates an act of inhumanity does not make it more inhumane than when a human does it, nor when a organ built of humans does it.

As I wrote about in my book, corporations are artificial persons, possessing artificial intelligence. That is, they have the means to dispatch pattern matching systems (humans, and lately computers) to carry out tasks on data and react based on the results of the information they receive. The corporation is one of many. Governments are artificial intelligences. NGOs, non-profits, departments, churches, all these organs, all have some level of artificial intelligence.

People of sufficient wealth and influence are themselves artificially intelligent. They can afford to hire people who pass off their work (ghostwriters, publicists, so on). And all of us have some level of social AI working on our behalf (I didn’t make this computer, this internet, my clothes, so on). But not nearly at the level the wealthy have, to augment their existence through the labors of others. An AI isn’t going to pull all the world out of a bit bucket. It will do like we already do and pull information from the rest of the world, ask people or machines to look up or research or build or whatever else. Just like corporations do today.

And what do corporations do with their artificial intelligence? Many of the same things great and scary, wonderful and terrible, that we worry AI will do. They abuse people, they educate people. They deliver aid to the needy, they create scarcity that makes people need aid. They help people move to a new country, to start a new life. They block people from moving to a new country, or deport them.

Churches have been around for thousands of years. Governments too. Sorry to break it to you, but AI already exists. It’s already a threat to peace and prosperity. It’s already hard to understand why it acts as it does, already hallucinates and operates in single-minded ways that ignore common sense.

Let us hope the next generation of artificial intelligence is a little smarter.

Monopoly Online

The Steam problem, but everywhere else too.

With some buzz about the American Innovation and Choice Online Act (AICOA) ( S. 2992: AICOA), it’s important to examine what systems like app stores, marketplaces, and search engines really do and look like.

My general position is favoring competition, but I think it’s as important to ask what competition looks like in a digital space as to ask why we don’t have it. We don’t have it, and the economy and people are hurt from not having it. Anti-competitive practices don’t help, but there are frictions that are also in play.

Online distribution

What a search engine like Google, a marketplace like Amazon, and an app store like Apple’s or Google’s all have in common is that they are middlemen. The customer wants to get something or get somewhere, and they can’t just get it or go there directly. They need assistance due to the enormity of the world. So Google lets them try to find information or a resource that fits their needs. Amazon lets them seek products matching their need. Apple app store and Google app store both let them find the software for their mobiles.

The bill’s attempt is modest. It makes it illegal to engage in stacking the deck. But it doesn’t, won’t promote real competition, which is a harder problem.

If you had competition, what would it look like? You want to search for banana peels, for your banana peel collection, so you go to different engines and search each one? I do that today. I go on duckduckgo and search, and if it doesn’t turn up what I need, and I think Google might, I’ll search on Google.

For a marketplace, same thing. You want to buy banana peels, so you go on Walmart and Amazon and see who has the best prices. Or if you find the prices on Amazon too high you try Walmart as a fallback.

With the app stores, it’s worse, of course. Because there’s limited options and each phone is doing the same things differently so that there’s not compatibility across Android and iOS.

But even if you stick to marketplaces (and assume you have accounts with each, no onboarding friction), it’s still a mess to find things across multiple stores, to make multiple purchases means multiple shipments. And if there are more than two big players, how many can people reasonably go through to buy banana peels?

Isn’t the problem that these middlemen exist as middlemen? That there isn’t a broader interface that cuts through all that friction, lets you see products?

If all competition means is that a bunch of new silly names enter the market that look mostly the same as the existing firms, that seems a waste. True competition will require different firms with different and limited objectives to each. Replacing the singular middlemen with mixtures of smaller businesses that cooperate.

To make that work, though, other changes would be needed. Amazon and Walmart have efficiencies of scale, they have warehouses that can combine ordered merchandise into single shipments, they have fluid logistics that move products around the country all day and night.

And there are environmental benefits to that. There are cost benefits to that. How can you replicate it without forcing oligopoly? How can a free system do the same thing?

Of course, there are frictions. A new store means a new login, entering your data, learning a new interface. Sellers have the same problem. Each new middleman means they have to replicate their process to another, slightly different platform. If they have dozens of products and dozens of middlemen, that’s a lot of work to keep things updated. If they have to fit in with shipping logistics, that’s a whole other set of processes that get replicated.

And it’s not like the existing options work that well, anyway. Amazon product listings are an ugly mess most of the time. Google can find some things, but it fails wildly on others. And the state of mobile operating systems, much less apps, leaves them looking like toys. Major newspaper apps don’t have a find-on-page function! Something every browser has had for decades!

This bill should pass, but it will only marginally improve the broken stupidity of the status quo.

A better system would require several changes, few to law, some to business structures, and most to software. An identity system would remove onboarding. Outside of a few members-only stores, you don’t have to fill out paperwork to buy something in stores. A vendor-agnostic logistics system would allow for product listings to be added and curated separately and without a particular middleman.

Products could be provisioned to warehouses and shipped based on similar functions that middlemen perform, but without the need for middlemen to act as gatekeepers. Returns would be handled in a similar way. This would create a class of independent warehousing and logistics firms who would follow industry standards to fulfill their functions.

The state of mobiles is probably the most difficult. Proper separation of OS from userspace, with all the security concerns inherent in software, is very difficult. Mobiles are kind of crappy anyway. I’m not really sure what it will take to fix this mess. There are all kinds of simple things one imagines using the mobile for, that is either impossible or requires a lot of separate services today. The gap between the science fiction ideal of a pocket computer and the thing you can buy today is as big as the difference between the lightning bug and the lightning. And that’s not something an antitrust bill does anything to address.

A Supply Chain-splainer

President Biden complained nobody had explained the supply chain, so I gave it a try.

First off, you might want to read “I, Pencil” (Wikisource: “I, Pencil”), an essay that’s quite popular with the libertarians. It gives a good idea how making anything is actually quite impossible. You see, to make anything would require making some thousand other things, which themselves would require making even more things. And that, children, is why we do not even exist!

Continue reading “A Supply Chain-splainer”