ChatGPT

[sighyoung]

Would love to hear from our academic friends on this one:

Northeastern college student demanded her tuition fees back after catching her professor using OpenAI’s ChatGPT

A senior at Northeastern University filed a formal complaint and demanded a tuition refund after discovering her professor was secretly using AI tools to generate notes. The professor later admitted to using several AI platforms and acknowledged the need for transparency. The incident highlights growing student concerns over professors using AI, a reversal of earlier concerns from professors worried that students would use the technology to cheat.

https://www.yahoo.com/news/northeastern ... 24481.html

I'm sure it happens, although part of me is surprised that this particular guy did it because if you've been teaching for years, you've got so much potential material to use that (aside from refreshing memory and doing additional research) you should be able to fill a sixty-minute to seventy-five minute class with substantive comments/activities without breaking a sweat. Hell, even if you don't know the reading all that well, you can still provide an hour's worth of decent information. And you don't have to put together lots of PowerPoint slides to it. In fact, why would you do it? I don't: I sometimes post my private lecture notes, but in class, I just focus on three or four key ideas about the writer and historical context, and then move on to the reading and class activities.

I could see people at the beginnings of their careers using AI to generate basic sets of notes, but I'd worry that the material would be incorrect. It takes a few years to develop enough teaching notes and in-class activities and class assignments that you feel comfortable teaching a specific course and you can save some time in course prep. I'm lucky my wife and I teach in the same period and teach the same kinds of courses: we've been able to share notes for years, and then put our own spin on them. But there are lots of resources available to help one put together a solid lesson plan.

However, adjunct faculty or people on one-year or three-year contracts are poorly paid while teaching eight classes across multiple campuses, and some professors are only rewarded for their research and could care less about teaching. Both will probably use AI to help them speed through course prep, feedback, and even grading.

Although it was really dumb of the professor to get caught (wouldn't you look at your own Powerpoint slides?), I'm not sure what the basis of the lawsuit is supposed to be. There was a typo on slide 3 and a person with seven fingers on his right hand in another slide? The teacher doesn't even have to do anything the student is complaining about to be said to have taught the class.

[sighyoung]

I will add that I've heard of teachers (not just college instructors, but grade school teachers as well) using ChatGPT to help generate comments on student papers, but even that is unnecessary.

My comments on essays and exams usually consist of two different kinds of feedback--those dealing with specific arguments and ideas, and technical feedback (formatting, titles, theses and thesis statements, topic sentences, grammar, spelling, etc.)

I definitely respond to unique properties in each paper (quality of ideas, quality of evidence, specific examples, and so on), but many problems appear in paper after paper (no thesis statement, insufficient evidence, lack of quotation or page citation, burying main ideas in the middle of paragraphs, poor transitions from one paragraph to the next.)

There are lots of ways of pointing out these more standard problems efficiently--a highlighted grading rubric or style sheet ("this is what an A paper does, this is what a B paper contains"), as well as re-using standard language in one paper after another ("provide more supporting evidence next time. Make sure your topic sentences state your paragraphs' main ideas more clearly.").

To me, there are different ways that teachers can clearly and efficiently provide this more repetitive kind of commentary, but I can see teachers turning to AI for boilerplate language.

Still, you need to show students that you have read and understood their writing, and provide feedback in such a way so that they avoid problems going forward. It isn't difficult, and really doesn't require use of AI to do it.

But again--I've developed a lot of these strategies over about 40 years of teaching. A newcomer might turn to AI for support, and the more problematic teachers probably do really on AI for grading. There have been teaching apps for a while that do some version of this that allow for human intervention and correction.

We risk, however, a cycle in which students rely more and more on AI to write (and to be fair, Microsoft Word and Grammarly already offer students lots of support when they write), and in which faculty use AI to grade. So GIGO.

[sighyoung]

I know I'm writing far too much about this, but let me put it this way: as a teacher, it is important not only to provide feedback, but to provide it in a way that is understandable and respectful. It is paramount that students know that I have respectfully read their work and that I genuinely care about them as learners. Lots of professors could say, "This paper is [expletive] smeared on a wall," which would convey information, but it's still bad teaching.

That Business professor failed to demonstrate that respect. It isn't difficult to do. At the same time, I don't know that he failed to impart concepts to his students. That would be the only legitimate way that that lawsuit could win.

But a lot of the authority I have in a classroom comes from my being seen as knowledgeable and respectful. You never want to do anything that loses that legitimacy or authority granted by students, and that professor lost it.

[mikechamp]

What is: unplug the server?

Oh, they have a battery backup? Ok. Soak them with a fire hose.

Problem solved.

The ‘godfather of AI’ reveals the only way humanity can survive superintelligent AI

Geoffrey Hinton, known as the “godfather of AI,” fears the technology he helped build could wipe out humanity — and “tech bros” are taking the wrong approach to stop it.

Hinton, a Nobel Prize-winning computer scientist and a former Google executive, has warned in the past that there is a 10% to 20% chance that AI wipes out humans. On Tuesday, he expressed doubts about how tech companies are trying to ensure humans remain “dominant” over “submissive” AI systems. “That’s not going to work. They’re going to be much smarter than us. They’re going to have all sorts of ways to get around that,” Hinton said at Ai4, an industry conference in Las Vegas.

In the future, Hinton warned, AI systems might be able to control humans just as easily as an adult can bribe 3-year-old with candy. This year has already seen examples of AI systems willing to deceive, cheat and steal to achieve their goals. For example, to avoid being replaced, one AI model tried to blackmail an engineer about an affair it learned about in an email.

Instead of forcing AI to submit to humans, Hinton presented an intriguing solution: building “maternal instincts” into AI models, so “they really care about people” even once the technology becomes more powerful and smarter than humans. AI systems “will very quickly develop two subgoals, if they’re smart: One is to stay alive… (and) the other subgoal is to get more control,” Hinton said. “There is good reason to believe that any kind of agentic AI will try to stay alive.”

Not everyone is on board with Hinton’s mother AI approach.

https://www.yahoo.com/news/articles/god ... 41649.html

[thrill]

Instead of forcing AI to submit to humans, Hinton presented an intriguing solution: building “maternal instincts” into AI models, so “they really care about people” even once the technology becomes more powerful and smarter than humans. AI systems “will very quickly develop two subgoals, if they’re smart: One is to stay alive… (and) the other subgoal is to get more control,” Hinton said. “There is good reason to believe that any kind of agentic AI will try to stay alive.”

There are like 700 sci fi books and movies about this.
This is the plot of iRobot. This is the plot of Murderbot (now streaming on Apple+). This is the plot of countless stories. Why are we acting like this isn't the most predictable, easy to understand [expletive] in the world? And yet, we're just letting these dumb Silicon Valley Icaruses (Icarusi?) fly straight in into the dang sun.

[sighyoung]

Yeah, Icaruses or Icari, I would assume.

And you're absolutely right about the large number of dystopian plots about intelligent machines.

But, add that to the dystopian plots about power-crazy scientists and money-crazy capitalists, and you can understand how we wind up with HAL in 2001: A Space Odyssey.

Why would machines have "maternal instincts" when those scientists and capitalists don't care about large swaths of humanity?

[thrill]

Sigh this brings me to a constant intrusive thought I have. Do capitalists watch the Alien franchise and root for Weiland Yutani? Do they watch Blade Runner and root for the Tyrell Corporation? Are they able to watch these movies, read these books (they aren't reading books), and root for the heroes and then go back to their daily lives of capitalist consumption without experiencing any cognitive dissonance whatsoever?

Does an ICE agent watch Star Wars and cheer when the stormtroopers come on screen? How is this circle squared by these people?

[GeddyWrox]

Sigh this brings me to a constant intrusive thought I have. Do capitalists watch the Alien franchise and root for Weiland Yutani? Do they watch Blade Runner and root for the Tyrell Corporation? Are they able to watch these movies, read these books (they aren't reading books), and root for the heroes and then go back to their daily lives of capitalist consumption without experiencing any cognitive dissonance whatsoever?

Does an ICE agent watch Star Wars and cheer when the stormtroopers come on screen? How is this circle squared by these people?

Excellent point/question.

[heyzeus]

Sigh this brings me to a constant intrusive thought I have. Do capitalists watch the Alien franchise and root for Weiland Yutani? Do they watch Blade Runner and root for the Tyrell Corporation? Are they able to watch these movies, read these books (they aren't reading books), and root for the heroes and then go back to their daily lives of capitalist consumption without experiencing any cognitive dissonance whatsoever?

Does an ICE agent watch Star Wars and cheer when the stormtroopers come on screen? How is this circle squared by these people?

They're the people who sing along to Rage Against the Machine, then get mad when Tom Morello speaks out against a rightwing political issue. These shows and songs are just fun words and stories to them. There's no cognitive dissonance because all fiction is just make believe.

[Arthur Dent]