12 Articles on Cheating with AI & AI Detectors

/

The Trouble With AI Writing Detection – Inside Higher Ed

College application season is here. So is the struggle to find out if AI wrote students’ essays – Cal Matters 

If using ChatGPT to write essays becomes widespread, those students who elect not to use it, who prefer to do the work themselves, may suffer a penalty for doing so. – Chronicle of Higher Ed

Results of a new survey flip the early narrative on ChatGPT—that students would rush to use it to cheat on assignments and that teachers would scramble to keep up—on its head. Half of students, ages 12-18, said they have never used ChatGPT. – Ed Week

OpenAI debates when to release its AI-generated image detector – Tech Crunch

Universities Rethink Using AI Writing Detectors to Vet Students’ Work – Bloomberg 

Identifying AI’s flaws motivates students and helps them build confidence, which can discourage cheating. Pointing out where it still really messes up is very powerful for empowering students to see their own strengths as human thinkers. – Chronicle of Higher Ed

Students cheat out of desperation so one professor will give multi-level assignments that force students to submit papers at various stages to keep track of their progress. – Yahoo News

The AI Detection Arms Race Is On And college students are developing the weapons, quickly building tools that identify AI-generated text—and tools to evade detection. – Wired

Simply leaving it up to students to decide whether they’re going to do the work, without further comment or intervention or negative sanction from me, is a failure of pedagogy. – Chronicle of Higher Ed

AI detectors have low efficiency, and simple modifications can allow even the most robust detectors to be easily bypassed. – Science Direct 

Suspicion, Cheating & Bans: AN Hits America's Schools (podcast) – New York Times

Status quo Bias

/

Should you stay or should you go? Status quo bias is our tendency to, when presented with a choice, prefer the current scenario as opposed to making a change. You can account for this natural bias by reversing the situation and the direction of change.

Status quo bias stems from a variety of human tendencies. A natural fear of change, our preference for familiarity, and laziness, all contribute. It's not our friend, either: Status quo bias contributes to many poorly thought decisions (like our tendency to overspend on big purchases).

Consider this: would you take a $13,000 wage increase to relocate to another city? Most people would say no. Yet consider the opposite: If you were living in another city, would you take a $13,000 wage decrease to move back to this one?

You can apply this reversal heuristic to smaller decisions, too. For example, instead of wondering whether you should spend a dollar for a chocolate bar, you could ask yourself whether you'd be willing to receive a dollar for skipping a chocolate bar for the day.

This quick reversal is a simpler version of the Reversal Test, a mental tool philosophers use to account for status quo bias.

Herbert Lui writing in LifeHacker

Problems are Signals

/

Americans in general have always admired growth. We admire the fastest growing companies and the cities that grew the most in the past decade. Magazines list the national economics that are growing the fastest. Bigger is better and bigger-faster is better still.

There is another kind of growth, which is much harder to measure. Its goal is not an increase in size (or intelligence or sophistication or experience or skill), but simply ripening. We overcome the barrier to growth as development when we are able to view our problems as signals that it is time to let go of the way in which we have been seeing and doing things and initiate a developmental transition.

The barriers to this kind of growth are overcome whenever we stop viewing our flaws and problems as things to be solved or removed and start viewing them as signals. What the problems are, really, are old solutions that have outlived their usefulness. From that point of view, whenever we do away with a problem instead of listening to its message, we trigger a string of events that lands us in trouble.

William Bridges, The Way of Transitions

10 Recent Articles about the Dangers of AI

/

The Bitter Lesson

/

There’s a famous essay in the field of machine learning known as “The Bitter Lesson,” which notes that decades of research prove that the best way to improve AI systems is not by trying to engineer intelligence but by simply throwing more computer power and data at the problem. The lesson is bitter because it shows that machine scale beats human curation. And the same might be true of the web. Read more at The Verge

The Tonic

/

Unselfish love for others is a tonic to the soul. At this very moment, someone needs you to care for him. Without you, his life may be incomplete. It could be a member of your family; possibly it is your neighbor or the person with whom you work. You see, you really do not have time to indulge in self-pity because of your past failures; already too many are stuck in the mud of self-pity, and they need you to lift them out of despair.  

Larry Kennedy, Down With Anxiety

15 Articles about The Potential Impact of AI on Medicine

/

15 Things people are trying to get AI to do–from solving crime to detecting wildfires

/

Can A.I. solve rape cases? To find out, a Cleveland professor programmed a computer to analyze thousands of police reports -Cleveland.com

Some in the (book) publishing world are already experimenting with AI programs in areas such as marketing, advertising, audiobook production and even writing, weighing their promise of supporting work done by humans against the threat that the machines ma. -NY Times

AI-powered technology may also help revitalize endangered languages, including by processing and storing languages and identifying language patterns. Additionally, AI may help accomplish these tasks at unprecedented speeds or just in time, before an endangered language goes extinct. -Inside Higher Ed 

Many in publishing are taking action to protect their work. The Authors Guild recently organized a petition signed by thousands of writers demanding that companies seek their approval before using their work to train A.I. programs. New York Times

Text With Jesus replicates an instant messaging platform, with biblical figures impersonated by the artificial intelligence program ChatGPT. The launching of the app stirred reactions ranging from amusement to accusations of blasphemy and heresy. -Religious News Service

Can ChatGPT become a content moderator? The technique is still not as effective as experienced human moderators, OpenAI found. But it outperforms moderators that have had light training.-Semafor

Can A.I. Detect Wildfires Faster Than Humans? California Is Trying to Find Out. -New York Times

AI providers begin to explore new terrain: chatbots in salary negotiations – Axios 

Coca-Cola launches beverage created with the help of artificial intelligence -Food Dive  

Get Ready for AI Chatbots That Do Your Boring Chores - Wired 

Alexa, will generative AI make you more useful? -Semafor  

Can AI predict, and try to prevent, homelessness? -NPR

ChatGPT was allegedly used to generate an apology statement about The Lord of the Rings: Gollum - TechRadar

Best Free & Paid AI Resume Builders: Build a Resume in Minutes - Tech.co

Multinationals turn to generative AI to manage supply chains - Financial Times

It’s All in the Attitude

/

Several years ago on an extremely hot day, a crew of men were working on the road bed of the railroad when they were interrupted by a slow moving train. The train ground to a stop and a window in the last car – which incidentally was custom make and air conditioned – was raised. A booming, friendly voice called out, “Dave, is that you?” Dave Anderson, the crew chief called back, “Sure is, Jim, and it’s really good to see you.” With that pleasant exchange, Dave Anderson was invited to join Jim Murphy, the president of the railroad, for a visit. For over an hour the men exchanged pleasantries and then shook hands warmly as the train pulled out.

Dave Anderson’s crew immediately surrounded him and to a man expressed astonishment that he knew Jim Murphy, the president of the railroad as a personal friend. Dave then explained that over 20 years earlier he and Jim Murphy had started to work for the railroad on the same day. One of the men, half-jokingly and half seriously asked Dave why he was still working out in the hot sun and Jim Murphy had gotten to be president. Rather wistfully, Dave explained, “twenty-three years ago I went to work for $1.75 an hour and Jim Murphy went to work for the railroad.”

Zig Ziglar, See You at the Top

We’re All Lousy Self-Evaluators

/

A stranger walks into a room and sits down behind a table. He picks up a piece of paper and read aloud a generic-sounding weather report. He completes his “report” in about 90 seconds and walks out of the room.

Next, you’re asked to guess his IQ.

You’re part of a psychological experiment, and you object to the absurdity of the request. I don’t know anything about that guy. He just came into a room and read a report. It wasn’t even his report- you gave it to him to read! How am I supposed to know his IQ?

Reluctantly, you make a wild guess. Separately, Fake Weatherman is asked to guess his own IQ. Who made a better guess?

Amazingly, you did, even though you know nothing about Fake Weatherman. Two (German) psychologists … conducted this experiment, and they found that the strangers’ IQ predictions were better than the predictions of those whose IQ was being predicted- about 66 percent more accurate.

To be clear, it’s not so much that you’re a brilliant predictor; it’s that he’s a lousy self-evaluator. We’re all lousy self-evaluators. College students do a superior job predicting the longevity of their roommates’ romantic relationships than their own.

Savor, for a moment, the preposterousness of these findings. Fake Weatherman has all the information, and you’ve got none. He’s got decades of data- year’s worth of grades, college entrance exams cores, job evaluations, and more. Fake Weatherman should be the worlds foremost expert on Fake Weatherman!

Chip & Dan Heath, Switch

10 Quotes Worth Reading about the Future of AI

/

Within five years everyone would have access to an AI personal assistant. He referred to this function as a personal chief-of-staff. In this vision, everybody will have access to an AI that knows you, is super smart, and understands your personal history. -Venture Beat 

Some experts in generative AI predict that as much as 90% of content on the internet could be artificially generated within a few years. -Bloomberg

Currently, most AI falls under narrow or specialized intelligence — good at one thing but pretty useless otherwise. However, we’re inching closer to Artificial General Intelligence (AGI), where machines can understand, learn, and apply knowledge across different domains. -Christophe Atten writing in Medium

It is certainly the case that many new technologies have led to bad outcomes – often the same technologies that have been otherwise enormously beneficial to our welfare. So it’s not that the mere existence of a moral panic means there is nothing to be concerned about.  But a moral panic is by its very nature irrational – it takes what may be a legitimate concern and inflates it into a level of hysteria that ironically makes it harder to confront actually serious concerns.  And wow do we have a full-blown moral panic about AI right now. -Marc Andreesen writing in a16z

All of the software we’ve ever used was engineered to work backward from an outcome. Its creators wanted to help you find a webpage or play a game or operate a laptop. Perhaps you’ve noticed that the major AI chatbots arrived with almost no user documentation or instructions. A lump of clay doesn’t come with instructions either. That’s what makes this moment unique — and so worthy of species-level #1 foam-finger pride. We humans have created a tool for potentially infinite tasks. Its imperfections are ours to solve — and its powers still ours to shape. -Washington Post 

“AI may cause a new Renaissance, perhaps a new phase of the Enlightenment,” Yann LeCun, one of the godfathers of modern artificial intelligence, suggested earlier this year. AI can already make some existing scientific processes faster and more efficient, but can it do more, by transforming the way science itself is done? Such transformations have happened before. -The Economist

DeepMind’s cofounder says generative AI is just a phase. What’s next is interactive AI: bots that can carry out tasks you set for them by calling on other software and other people to get stuff done. “Technology is going to be animated. It’s going to have the potential freedom, if you give it, to take actions. It’s truly a step change in the history of our species that we’re creating tools that have this kind of, you know, agency.” -MIT Tech Review

What If the Robots Were Very Nice While They Took Over the World? First it was chess and Go. Now AI can beat us at Diplomacy, the most human of board games. The way it wins offers hope that maybe AI will be a delight. -Wired

People need to develop “rugged flexibility,” to manage change most effectively. In other words, people need to learn how to be strong and hold on to what is most useful but also to bend and adapt to change by embracing what is new. -Venture Beat

Imagine if your brain got 10 times smarter every year over the past decade, and you were on pace for more 10x compounding increases in intelligence over at least the next five. Throw in precise recall of everything you’ve ever learned and the ability to synthesize all those materials instantly in any language. You wouldn’t be just the smartest person to have ever lived — you’d be all the smartest people to have ever lived. (Though not the wisest.) That’s a plausible trajectory of the largest AI models. -Washington Post 

We seem to be in what I can only call an “AI lull.” The initial excitement about ChatGPT, which started in January, has receded. Do not be deceived. While the hype and marketing may have died down, at least on the retail side, the AI revolution will continue. -Bloomberg

What Overinvolved parenting does to Kids

/

When parents have tended to do the stuff of life for kids—the waking up, the transporting, the reminding about deadlines and obligations, the bill-paying, the question-asking, the decision-making, the responsibility-taking, the talking to strangers, and the confronting of authorities, kids may be in for quite a shock when parents turn them loose in the world of college or work. They will experience setbacks, which will feel to them like failure. Lurking beneath the problem of whatever thing needs to be handled is the student’s inability to differentiate the self from the parent.

When seemingly perfectly healthy but overparented kids get to college and have trouble coping with the various new situations they might encounter—a roommate who has a different sense of “clean,” a professor who wants a revision to the paper but won’t say specifically what is “wrong,” a friend who isn’t being so friendly anymore, a choice between doing a summer seminar or service project but not both—they can have real difficulty knowing how to handle the disagreement, the uncertainty, the hurt feelings, or the decision-making process. This inability to cope—to sit with some discomfort, think about options, talk it through with someone, make a decision—can become a problem unto itself.

Julie Lythcott-Haims, How to Raise an Adult: Break Free of the Overparenting Trap and Prepare Your Kid for Success