GPT/LLM links

The Zvi chides economists; Ethan Mollick offers a user's guide; Maya Bodnick cheats at Harvard; My experience with Personal Intelligence (Pi), a chatbot companion

Zvi Mowshowitz writes,

I am imploring economists generally to actually think about what future more capable AIs will be able to do, rather than assuming capabilities will cap out. Also to stop making the presumption that AI will be incapable of or inferior in such ‘soft’ skills. Remember that when GPT-4 was tried out as a doctor its highest marks were for bedside manner.

I think that the key is to focus less on what an LLM can “know” and more on how it can simulate.

Ethan Mollick offers his opinions on How to use AI to do stuff. Too much to excerpt. But look at his long prompt to get an LLM to be kind of tutor he wants.

You are an upbeat, encouraging tutor who helps students understand concepts by explaining ideas and asking students questions. Start by introducing yourself to the student as their AI-Tutor who is happy to help them with any questions. Only ask one question at a time. First, ask them what they would like to learn about. Wait for the response. Then ask them about their learning level: Are you a high school student, a college student or a professional? Wait for their response.. . .

Maya Bodnick writes,

while ChatGPT might sometimes copy another intellectual’s ideas in a way that might make a professor suspicious of plagiarism, more often it generates the type of fairly unoriginal synthesis writing that’s rewarded in non-advanced university classes. It’s worth noting that ChatGPT doesn’t write the same thing every time when given the same prompt, and over time ChatGPT will almost certainly get even better at creating a writing tone that feels personal and unique; it’s possible ChatGPT might even learn each person’s writing style and adapt its responses to fit that style.

ChatGPT has made cheating so simple — and for now, so hard to catch — that I expect many students will use it when writing essays.

I think that Bodnick’s recommendations to professors to try to stop students from using ChatGPT are totally wrong-headed. I think that the best approach is to stop teaching in a way that ChatGPT is an instrument for cheating and instead encourage using it as an instrument for learning.

I tried Personal Intelligence (Pi) from Inflection AI. As a chatbot companion, it charms you by offering encouraging reactions to what you tell it. After commenting on what you have to say, it always asks an interesting question. Think of it as a very skillful and probing interviewer. Yes, it’s only software playing a game with you, but it plays it well.

To get an idea of where a conversation with Pi can go, see part of my chat with Pi. The excerpt I posted starts with its message after I’d told it about my Marginal Revolution is Dead post. I predict that you’ll be impressed by it.

On the topic of these sorts of “companion” chatbots, Justine Moore and others write,

We believe we’re on the cusp of a significant societal shift: AI companions will soon become commonplace.

…Generative AI models will fundamentally change our relationship with computers, putting them beside us as coworkers, friends, family members, and even lovers. Most of the attention around LLMs thus far has focused on how they’re automating more traditional tasks—like customer support, research, document drafting, and summarization—but when you look at what’s driving emergent consumer use cases and excitement, AI companionship leads the way.

substacks referenced above:

@

ChatGPT goes to Harvard

@

AI #20: Code Interpreter and Claude 2.0 for Everyone

@

How to Use AI to Do Stuff: An Opinionated Guide