Geek Stuff category archive

Under the pretext of a quibble over terminology, psychology professor Gregg Henriques takes a deep dive into why and how AI Chatbots and LLMs get so much so wrong so often. Here’s a tiny bit from his article (emphasis added):

For example, when my family was playing around with ChatGPT, we wanted to see if it “knew” who my father was. My dad, Dr. Peter R. Henriques, is a retired professor of history who has written several books on George Washington. ChatGPT respond correctly that my dad was a biographer of Washington; however, it also claimed, wrongly, that he wrote a biography on Henry Clay. This is an example of a hallucination.

Where do hallucinations like these come from? LLMs like ChatGPT are a type of artificial intelligence that run algorithms that decode content on massive data sets to make predictions about text to generate content. Although the results are often remarkable, it also is the case that LLMs do not really understand the material, at least not like a normal person understand things. This should not surprise us. After all, it is not a person, but a computer that is running a complicated statistical program.

At Psychology Today Blogs, Joyce Marter debunks de bunk. A snippet:

While multitasking may seem like a productivity booster, it can also lead to decreased focus, poorer work quality, and increased stress levels. Multitasking has been proven to reduce productivity and job performance . . . .

Follow the link for context.

The Register reports on a New York law firm that tried to use ChatGPT to justify a ginormous billing.

The judge was not impressed.

Mageia v. 9 with the Plasma desktop. Firefox is shaded near the top of the screen under the Plasma menu. Xclock is in the upper right, GKrellM in the lower right. The wallpaper is from my collection.

Recently, I ran an online upgrade from v. 8 to v. 9. The online upgrade from v. 7 to v. 8 went smooth as glass, and this one seemed to also, but, when it was done, I was unable to run updates or install new software from the repos. I poked at the problem for a while, but was unable to resolve it, so last night I installed v. 9 from optical media while listening to a BBC Lord Peter Wimsey mystery at the Old Time Radio Theater.

The installation went quickly and easily, and, as I have a separate /home partition, when I fired it up, all my configuration files were still in place without my having to restore anything from backup (and, no, you can’t do that on Windows). I’m currently cleaning up the dust bunnies, such as, for example, installing the few applications, like Xclock and GKrellM, which are not part of the standard Mageia installation.

Thanks to the most excellent detective work of my hosting provider’s tech support staff, it has been deemed necessary to remove the NOAA Weather plugin that was on the sidebar over there —–> for years, as the plugin has not been updated for several years and is no longer compatible with the most recent versions of php, the scripting language that powers WordPress, which in turn powers this geyser of genius. (Or is it a drenching of drivel? Inquiring minds want to know.)

As one who wore a headset for over half a decade, I must say that my hosting provider’s tech support staff is superb.

I know, because I’ve been there.

At Psychology Today Blogs, Dr. Marlynn Wei tales a look at the psychological implications of the spread of deepfakes and “AI” clones and lists half a dozen dangers. Here’s one (emphasis in the original):

4. Creation of false memories

Research in deepfakes shows that people’s opinions can be swayed by interactions with a digital replica, even when they know it is not the real person. This can create “false” memories of someone. Negative false memories could harm the reputation of the portrayed person. Positive false memories can have complicated and unexpected interpersonal effects as well. Interacting with one’s own AI clone could also result in false memories.

Her article is a worthwhile read, and prends garde a toi.

Via Bruce Schneir, here’s a study that demonstrates that AI can be made more human-like.

That is, it can be “trained” to deceive.

As the song* says

The things that you’re liable
To see in your large language model,
They ain’t necessarily so.

_________________

*With apologies to George Gershwin.

At Psychology Today Blogs, Bill Sullivan offers yet more evidence that “social” media isn’t, particular for the young. Here’s a bit of his article:

Studies have shown that social media use, especially among younger people, increases materialism—a drive to accumulate and flaunt money and possessions. Such a mindset implies that external goods like Italian sports cars, Cartier watches, or gourmet dining are signs of a person’s status. Materialists gauge their own worth—and the worth of other people—by how much money and bling they’ve acquired, rather than judging internal goods like character. The materialistic mindset fosters attitudes of competition, envy, and greed.

Madonna’s claim that we are living in a material world is backed by convincing data.

Follow the link, then be sure to post it to your Zuckerborg or Muskrat page.

Under-valuation frolics.

Just because you see–er–hear it on a computer–er–device, it ain’t necessarily so.

The Second Circuit has referred attorney Jae S. Lee to a grievance panel for citing a fake case made up by ChatGPT in a complaint and never checking to see if the case spit out by the AI was real.

Details at the link.

A Consumer Reports study details the extent to which you have been assimilated by the Zuckerborg and its enablers. Indeed, perhaps the most astounding bit in the report is the number of Zuckerborg enablers who “share” your data with Facebook.

Follow the link for the article.

Using a panel of 709 volunteers who shared archives of their Facebook data, Consumer Reports found that a total of 186,892 companies sent data about them to the social network. On average, each participant in the study had their data sent to Facebook by 2,230 companies. That number varied significantly, with some panelists’ data listing over 7,000 companies providing their data.

One more time, “social” media isn’t.

You don’t use “social” media. It uses you.

H/T Bruce Schneier for the head up.