Category: Geek stuff

If it involves physics, math, computer science, electronics or has general geek cachet, this is the place for it.

AI image generation and the credibility of photos

When AI-assisted photo manipulation is easy to do and hard to detect, the credibility of photos as evidence is diminished:

No one on Earth today has ever lived in a world where photographs were not the linchpin of social consensus — for as long as any of us has been here, photographs proved something happened. Consider all the ways in which the assumed veracity of a photograph has, previously, validated the truth of your experiences. The preexisting ding in the fender of your rental car. The leak in your ceiling. The arrival of a package. An actual, non-AI-generated cockroach in your takeout. When wildfires encroach upon your residential neighborhood, how do you communicate to friends and acquaintances the thickness of the smoke outside?

…

For the most part, the average image created by these AI tools will, in and of itself, be pretty harmless — an extra tree in a backdrop, an alligator in a pizzeria, a silly costume interposed over a cat. In aggregate, the deluge upends how we treat the concept of the photo entirely, and that in itself has tremendous repercussions. Consider, for instance, that the last decade has seen extraordinary social upheaval in the United States sparked by grainy videos of police brutality. Where the authorities obscured or concealed reality, these videos told the truth.

Perhaps we will see a backlash against the trend where every camera is also a computer that tweaks the image to ‘improve’ it. For example, there could be cameras that generate a hash from the unedited image and retains it, allowing any subsequent manipulation to be identified.

Related:

Cleese for the record

But in other areas I was becoming less diffident—or, in St. Peter’s parlance, less “wet.” Indeed, on one occassion, I actually got into a fight with a boy who was teasing me. There I was, lying on the floor, grappling with him, like a proper schoolboy; I even banged his head on the floor, at which point I thought, “Oh my God! If I start losing, he’ll do this to me,” and then, of course, started losing. Fortunately my form master, Mr. Howdle, arrived and broke the fight up. Funnily enough, it was about then that the bullying stopped. This first fight also proved to be my last. I had thought so, anyway, until I read in the Sunday Times recently that I had a fight with Terry Gilliam in the ’80s. I think this is unlikely: owing to the relatively rare occurrence of fisticuffs in the Cleese life it must be statistically probable that I would remember such uncommon events; they would tend to stand out sharply from the rather less pugilistic tone of the rest of my life. And I definitely don’t recall having a fight with Terry Gilliam. May I also point out that if I had, I would almost certainly have killed him. I think the only possible explanation for the Sunday Times article—if it was true—was that Terry attacked me, but that I failed to notice he was doing so. Terry is very short, due to his bandy legs, so when he scuttles around, he stays so close to the floor that it can be difficult to see what he is up to down there.

Cleese, John. So, Anyway… Penguin Random House, 2014. p. 43 (italics in original)

Enjoying Toronto’s Bike Share in the summer

On Wednesday evening, I did a 55km bike ride: east from the U of T campus across the Don into the beaches area, down to the southern tip of Tommy Thomson Park, then along the waterfront for a picnic dinner at a Queen’s Quay grocery store, and up the hill to The Perch.

These animations show the ride in yellow as well as all my previous walks and rides since 2020 in green:

Taleb on the domain dependence of knowledge

I used to attend a health club in the middle of the day and chat with an interesting Eastern European fellow with two Ph.D. degrees, one in physics (statistical no less), the other in finance. He worked for a trading house and was obsessed with the anecdotal aspects of the markets. He once asked me doggedly what I thought the stock market would do that day. Clearly I gave him a social answer of the kind “I don’t know, perhaps lower”-quite possibly the opposite answer to what I would have given him had he asked me an hour earlier. The next day he showed great alarm upon seeing me. He went on and on discussing my credibility and wondering how I could be so wrong in my “predictions,” since the market went up subsequently. Now, if I went to the phone and called him and disguised my voice and said, “Hello, this is Doktorr Talebski from the Academy of Lodz and I have an interrresting prrroblem,” then presented the issue as a statistical puzzle, he would laugh at me. “Doktorr Talevski, did you get your degree in a fortune cookie?” Why is it so?

Clearly there are two problems. First, the quant did not use his statistical brain when making the inference, but a different one. Second, he made the mistake of overstating the importance of small samples (in this case just one single observation, the worst possible inferential mistake a person can make). Mathematicians tend to make egregious mathematical mistakes outside of their theoretical habitat. When Tversky and Kahneman sampled mathematical psychologists, some of whom were authors of statistical textbooks, they were puzzled by their errors. “Respondents put too much confidence in the result of small samples and their statistical judgments showed little sensitivity to sample size.” The puzzling aspect is that not only should they have known better, “they did know better.” And yet…

Taleb, Nassim Nicholas. The Black Swan: The Impact of the Highly Improbable. Random House, 2007. p. 194-5 (italics in original)