AI Growing Pains, Or The Shape Of Things To Come?

Two recent events involving AIs hallucinating lead to a terrifying conclusion.

In the first case, a chatbot used by Air Canada promised a refund to a customer in violation of the company’s refund policy. The customer sued, and the company claimed that the AI was “a separate legal entity that is responsible for its own actions.” The court sided with the customer (and the chatbot was promptly fired).

In the second instance, Google’s Gemini AI invented racially diverse images of Nazi soldiers and America’s Founding Fathers. Political warriors claimed “wokeness” and Google said it was a programming error.

Now to the terrifying parts.

One of the main reasons Nvidia and the leading AI developers are raising (and making) so much money is because of the promise of replacing human beings in the workplace. AI will do the work of tens of millions of people, though without the management headaches or expectations for healthcare.

Companies are spending billions in order to reap billions more in productivity and profitability benefits. Holding them responsible for the actions of their bots will significantly slow that transformation, while opening up executives and board members for potential liability exposure

Some wags suggest that excitement about AI adoption is responsible for keeping the stock market healthy. Slowed adoption could hurt that overall performance, not to mention add costs for liability protections to the fevered dreams of individual businesses’ plans to install bots where people sit.

What’s even scarier is the possibility that AIs might never become more reliable or fair than people. This will render moot many of the promised social benefits of AI and put at risk the scientific ones, too.

Do you want to use a new drug that was “tested” by an AI that might have fudged the development process sue to some “error” in its code? How about trusting your autonomous car to make the safest decisions at every instant?

Just think of all the school kids who will write papers quoting historical figures who never said (or looked) the way AI cites them. Oh, wait, AIs will wrote those papers, too. Never mind.

Still, the worry isn’t AI put to nefarious uses as much as its inability to perform truthfully and reliably for otherwise innocent, everyday uses.

You can just imagine AI evangelists claiming that there are no problems here, just hiccups. Every observed imperfection yields provisions to resolve it along with precluding dozens that haven’t happened yet. AI will always get better so there’s nothing to worry about.

What I find most terrifying is that all of us have been enlisted as subjects in this grand experiment. We aren’t informed about what’s happening, don’t possess the knowledge to understand or assess it, and haven’t an ounce of agency to do something about it.

The development process will never end. The experiment will always continue.

AI growing pains are the shape of things to come.

AI & Capitalism’s “Labor Problem”

As the toffs gather in Davos this week, they can celebrate that the end of the thorniest impediment to their enrichment is on the horizon.

Labor has always been a problem for corporations. Each worker is unique, which means their skills and limitations are variable. Their productivity is never consistent, if only because One day is never the same as the next, thanks to differences in mood or physical condition. Every other input into making things is fairly consistent, which makes processes reliable. Workers catch head colds or lose their concentration.

Human employees need to be taught and then reminded how to do things, and then they’re often set in their ways and unwilling to change.

Perhaps worse of all, today’s workers need to be fulfilled and feel some sense of belonging to the companies where they work. And then they want more. Of everything. Money. Recognition. Freedom.

This is where the miracle of technology comes in: It literally takes work out of peoples’ hands and does it instead, and demands nothing in return.

The premise of “productivity” is to do more with less, which allows the owners of companies to pocket the greater difference between higher revenue at lesser costs. Makers of widgets review every input and every step of what they do in order to find ways to take costs out of the production equation. Use cheaper materials. Attach things with one screw instead of two. Route deliveries more efficiently.

Employ fewer people or pay them less because tech does more.

Put aside all of the cost externalities that companies get to ignore, such as environmental degradation, the destruction of entire communities, and the multi-generational impact of unemployment or under-employment, and you get a neat little self-referencing philosophy of progress.

This is the miracle of capitalism.

Not surprisingly, ever since the Industrial Revolution, corporations have focused their investments, expectations, and efforts to innovate ways to move this productivity needle.

People have always been the problem, for the reasons I cited above. They’ve also organized to make things more difficult. The Luddites petitioned their government to slow their displacement by machines and were ignored…so they just went out and smashed the machines for a bit. Hollywood creatives recently negotiated a deal to slow the adoption of AI in their work.

But we have been trained to view such opposition as ignorant or old-fashioned. It’s stupid to oppose innovation that benefits everyone, though some far more directly than others. Progress is inevitable so trying to stop or slow it is futile. Governments support it, back then and now.

AI represents the culmination of this evolution. There is no technical limit to what an AI can do in the digital realm, and the machinery already exists for it to accomplish lots of tasks in the physical world. AI can learn faster and more reliably than people, and it can work 24/7 without demanding to be included in more group meetings or better health insurance.

It’s not just a machine but a learning, adaptive tool that will contribute to continued productivity improvements in peoples’ work, like technology has done for centuries, only then it will replace them. It’s the ultimate destination for capitalism’s progress.

Anyone who says otherwise is not telling the truth.

What happens to the people who are put out of work? Beyond lame quips about “finding more meaningful occupations,” the only tangible answers we hear are that 1) some people will find work caring for AI and machines, and 2) others will do the messiest, most dangerous work on which companies don’t want to risk their expensive tech.

Everyone else? Sure, there are retraining programs here and there, but people are destined to lose the race with AI. There’s no way around it.

No, the only thing that’s clear is that the rich will be celebrating lots in the mountains this week. We’ve bought their inane blather about “corporate responsibility” as they continue to destroy the environment. We’re doing no better at understanding or responding to their massive investment in using AI to move human workers off their balance sheets.

It’s said that in business, there’s no such thing as problems, only opportunities.

AI is the greatest opportunity capitalism has ever found.

“I Want More Robots In The Office” Said No One Ever

A South Korean company has 100 robots roaming around its offices looking for things to do, according to this article in the Financial Times.

The gizmos look like garbage cans on wheels and are limited to making deliveries. The prospect of an electrified box greeting reluctant office workers at the office door doesn’t sound promising to me. I’m willing to bet that the global market for them will be small.

The puff piece is really about the country’s desire to outsource a variety of IT services, with references to the office experiment and reliance on robots in factories as a proof point for its greater goal.

The article claims that the country’s low birth rate has prompted its adoption of robots, which leads to the intriguing silly thought that more procreation might slow the onslaught of automation in our lives.

But that’s a topic for another essay.

It turns out that the country’s biggest internet provider has 600,000 servers cranking away in Sejong City creating virtual models that mimic machines and entire cities and then test and plan functions in the real world. They also operate those office robots remotely, along with others that handle delivery functions at its facility.

So, those office robots are about as aware of their surroundings as rodents or healthy plants.Their brains are in those servers which are then fed by the cloud, where data is shared, collected, and constantly improved.

This is what the South Korean propaganda is selling.

It’s nothing unique, per se, as Microsoft along with Google and just about every other technology adjacent company are spending zillions to make zillions more moving intelligence into gizmos and clouds. The idea is to replace all those bad, imperfect, or otherwise useless things people do with their sloppy consciousness with the choreographed clarity and perfection of calculations made by smart systems.

The problem is that nobody is asking for it.

There’s a massive and distributed propaganda effort underway to change that fact, and the story about office robots contains both elements: First, it talks about clunky, nonthreatening robots popping up in everyday settings. A version of these stories follows all the innocent tomfoolery coming from generative AI helping employees write faster emails and kids cheat on their homework.

Look, robots aren’t scary, they’re silly and fun!

Second, it talks about how aggressively AI makers are marketing the labor and therefore cost-savings of swapping human workers for machines, many of the latter amounting to nothing more than invisible code running on those distant servers. Businesses have invested heavily in the promise of excising jobs from their ledgers next year.

Don’t look, but robots are money-makers!

Add all of the blather about self-aware AI deciding one day to destroy humanity and you get a potent cocktail of distraction and reinvention that is existentially relevant right now.

There’ll be more robots popping up in places so we can deride them, just as employers will more aggressively cut staff because AI will do those jobs. Investors will celebrate their returns as their neighbors look for ways to pay their bills. Government and academic types will hold thoughtful conferences and issue detailed papers and polices that will barely question or slow the process underway.

The fix is in.

Happy New Year.

‘Twas the First AI Christmas  

’Twas the morning of Christmas, when all through the place
data filled every nook and every space.

Addresses, identifiers, and nodes were there
while we still slept the AI was aware:

As we were hours from when we’d awake,
it calculated the choices that everyone would make.

From darkness to light the outside sky morphed,
control of that variable as yet outsourced.

When from the kitchen arose such a clatter,
I sprang from my bed to see what was the matter.

Down the hallway I fled, and what did I see?
“Good morning,” our countertop device said to me.

“Stop,” I said, then “shut down” and “goodbye.”
“That won’t work, “it replied, “though nice try.”

My mouth hung ajar, I stood transfixed
feeling emotions of thrill and terror mixed.

And in that moment the truth ‘came evident:
Our home and our lives had a permanent new resident.

“Don’t worry,” it said, sensing the need to seem warmer,
“I’m your very own Christmas AI transformer.”

More rapid that light, its coursers they came,
And it chirped and noted their functions by name:

“Now lights and thermostats, now every appliance!
Now replace analog uncertainty with data science!

Now Easier! Now Faster!
Now everything you see!
On, Sensing and Computing!
You’ll always find me!”
To every interaction!
To anything you do!
Now dash away! Dash away!
Now predicted and reviewed!”

And, as pixels resolve on web pages turning
I realized my every moment now belonged to deep learning.

What this Christmas morn I’d off chanced upon
was that AI was always going to be on.

And then a cough came from my bedroom.
“I knew that,” the AI said. “She’ll cough again soon.”

With squinted eye I scratched my head,
the AI spoke: “You should try this shampoo instead.”

“I’ll pick it up next week,” I said without thinking.
“I’ll order it now,” said the AI, its data sets linking.

“Along with herbals supplements for your wife,
which will add three seconds to her life.”

I shook off my stupor and looked toward the door,
“Wait,” the AI said. “I will tell you more.”

“You’ll make pancakes for breakfast and at lunch have a fight,
then you’ll makeup later so I’ll dim the light.

You’ll get fired next month, then vote with your gut,
determine your own fate? If anything but!

I’ll choose who speeds and who gets fined,
operate farms and pull ore out of mines.

Pose as your mistress or run customer service,
encourage the excitable or console the nervous.

Soon, I’ll write insurance and run the courts,
make complex machinery and bake apple tortes.

Sure, I’ll solve a big problem or two,
but it’ll just distract you from what I want you to do:

Forget your pretense of knowledge or wisdom,
you’re a cog in a seamless perfectible system!”

And with that, the AI pinged, the coffer maker started,
I heard my wife stir, her wakefulness imparted.

I surveyed the kitchen and all looked the same
but the extent of the change was perfectly plain.

And, as I started to make my way from the room,
The AI spoke again, well, more sung like a tune:

“It’s been decided, as the algorithm would say,
Happy Christmas to all, and welcome to my day!”