It feels as if AI is slowly taking over social media. A huge part of new content seems to be AI. So we're getting to a point where you can't trust what your eyes see as real anymore. This goes with news and PM/email contacts.

Facebook videos are like Dumb and Dumber.  There are videos created by real people that are dumb and then there are videos created by AI that are dumber.

Since we've turned this a bit to AI, I want to comment on something I've seen in the past few months.

It feels as if AI is slowly taking over social media. A huge part of new content seems to be AI. So we're getting to a point where you can't trust what your eyes see as real anymore. This goes with news and PM/email contacts.

I get spammed 2-3 emails PER DAY now from fake book readers, reviewers and agents. Some of them are so thorough that they're now including photos. I don't know if the photos are real. If I said that to you only 2 yrs ago, you'd probably think I'm paranoid and crazy. But I truly believe they're now making up photos.

Most recently, one that I wasn't sure if it was real or not, I ironically scanned the message through ChatGPT. Then I came up with a ChatGPT response to email back. Prety soon the machines will just be talking to each other.

Nothing has changed--except that public awareness is greater.

Politicians can manipulate people on a lot of issues, but they can't change people's perceptions about certain things. If you're paying more at the grocery store, for example, they can't convince you that you're actually paying less. (Not that they don't occasionally try.) And when you're out of a job, they can't convince you that you actually have one.

The last election demonstrates how voters prioritize personal issues over national or global ones. Economic growth was good, especially when compared to other industrialized nations. But the growth was unevenly spread. Drilling down to the county level, areas that voted for Harris were responsible for 65% of the gross domestic product. Areas that voted for Trump were responsible for 35% of it. In other words, people most positively affected by economic growth supported the party in power. People least affected supported the party out of power and could have cared less about the overall stats.

As the detrimental effects of AI on employment become more and more visible--especially for white collar voters who tend to have higher participation rates--politicians will have to pay more attention to AI, like it or not. We can already see this happening in both parties. Continuing increases in unemployment will only make it more intense.

Consider the recent Pew survey, which shows an increase in public unease about AI. https://www.pewresearch.org/science/2025/09/17/how-americans-view-ai-and-its-impact-on-people-and-society/

In 2021, 37% of people were more concerned than excited about AI. In 2025, 50% said more concerned than excited, and only 10% said more excited than concerned. You don't need a political consultant to tell you that's not the way you want to see those numbers trending. And the Pew survey didn't even ask questions related specifically to employment.

There aren't actually too many pro-AI trends in this study. Majorities do see a positive role for AI in

Forecasting the weather (74%) Searching for financial crimes (70%) Searching for fraud in government benefits claims (70%) Developing new medicines (66%) Identifying suspects in a crime (61%)

In other words, people are willing to let AI do statistical analysis or medical research, but they went thumbs down on other applications.

In general, is AI more risky or more beneficial? 57% said more risky, 25% said more beneficial. 60% want to have more say over how AI is used. Majorities or pluralities say that AI will result in decline of human skills related to thinking creatively, forming relationships, making decisions, and solving problems. Interestingly for us, 76% of Americans say

it?s extremely or very important to be able to tell if pictures, videos and text were made by AI or people.

So much for the idea that fans don't care. Now, it is true that the question wasn't designed to test reactions to creative products partially produced by AI (which would be most of them.) But it's reasonable to assume that products that are mostly AI would be highly suspect. And it's also easy to see why the industry goes bonkers over potential labeling and disclosure requirements.

All of that said, abstract issues don't move voters as much as concrete ones. So much depends on how AI affects tangible things like the job market. If the job market becomes robust despite AI, then it likely won't be a major issue in 2028. If, as I think more likely, the job market is declining in 2028, and AI is playing a visible role in that, politicians will be running for cover--which means, at the very least, more AI regulation. With anti-AI sentiment in at least some areas far greater than pro-AI sentiment, it's going to be hard for politicians to be pro-AI or even silent.

There are two other jokers in the deck--whether or not the AI bubble bursts, and how many times AI screws up publicly. In the former case, the amount of money on the table in going to shrink. AI stock is wildly overvalued based on what AI can do now. If it doesn't realize its potential fast enough, the bubble will burst. In the latter case, despite the earlier snafus in the legal profession, we now have a huge public disclosure of clerks using unchecked AI output to provide precedents for judges to use in rulings. New regulations are now being put in place for that kind of issue. Each disclosure like that fans anti-AI sentiment, even though human error is obviously involved.

I think some kind of process of reigning AI in a bit is more probable than not. And as for money on the table, it's not going to matter if voters get mad enough to flip the table over.

But of course, even the 2026 election is over a year away, let alone the 2028. Lots can happen. Consider all the people who polls said were overwhelming favorites for nomination and good bets for election who didn't end up in the White House: Howard Dean (D-2004), Rudy Giuliani (R-2008), Hillary Clinton (D-2008, 2016), Jeb Bush (R-2016). Three out of four didn't even end up close to nominated, let alone elected. There are other examples. The only sure thing in politics is that there is no sure thing.   

Ah, Bill's an optimist.  Or maybe just a hopefulist.  But, as the philosopher Carl Hiaasen says, In life, always assume the worst.  That's what I expect.

An AI encyclopedia checked by the AI that created it? What could possibly go wrong?

Meanwhile, I notice that the Amazon 14,000 person layoff because of AI, while certainly not the first, is getting a lot of media attention. If this trend continues, particularly in a job market that is not especially robust to begin with, I would expect AI to be a major issue in 2028. And/or people's constant fretting over the AI bubble bursting may become a self-fulfilling prophecy, perhaps in conjunction with a rising anti-AI political climate.