Saturday, April 29, 2023

Weekly Sporto bookmarks (weekly)

  • tags: chatGPT AI tools

    • Being “good at prompting” is a temporary state of affairs. The current AI systems are already very good at figuring out your intent, and they are getting better. Prompting is not going to be that important for that much longer. In fact, it already isn’t in GPT-4 and Bing.
    • By breaking the pattern, you can get much more useful and interesting outputs. The easiest way to do that is to provide context and constraints.
      • Be creative/make any assumptions you need. This will tend to remove some of the constraints of practicality around AI answers, and can be useful if you are trying to generate something novel.

      • Show your work/provide sources/go step-by-step. The AI will make up information that it does not have access to. There is some evidence that asking it to show its work, or its sources, reduces that risk somewhat. Even if it doesn’t, it can make checking work easier.

      • Write me code and tell me how to use it. If you can’t code, you might be able to now. AI can do some amazing things with Python programs, and tell you exactly how to run it. I don’t know coding, but I have written a dozen Python programs in the last month. If there are errors in the code, and there likely will be, just give them to the AI to correct.

      • Write a draft/provide an example. If the AI refuses to do something (“you should be creative and write your own novel, I can’t help”, sometimes asking it to provide something like a draft can get it to produce results.

    • Another bit of sorcery: After it gives you an answer, ask it to critique its own response, poke holes in it, then ask it to improve its response based on that critique.
  • tags: chatGPT AI tools Linkedin

  • tags: chatGPT AI tools

    • More elaborate and specific prompts work better.
    • But as a tool to jumpstart your own writing, multiply your productivity, and to help overcome the inertia associated with staring at a blank page, it is amazing.
  • "You should make sure you are forcing Bing to look something up with every query. Things that have worked for me include prompts like First research ____. Then do ____ or else prompts like Look up ____ on Reddit/in academic papers/in the news. Then use that to ____. Either way, you want to trigger the “searching for” label to get good results. "

    tags: chatGPT AI tools

    • You should make sure you are forcing Bing to look something up with every query. Things that have worked for me include prompts like First research ____. Then do ____ or else prompts like Look up ____ on Reddit/in academic papers/in the news. Then use that to ____. Either way, you want to trigger the “searching for” label to get good results.
    • Indeed, Bing is at its most powerful, and most different from ChatGPT, when it is looking up data and connecting diverse sets of information together. It is often a startling good data analyst, marketer, and general business companion.
    • One trick for using its power is to ask it for charts that pull together lots of information. Analyze the market for alternative milk products. provide a chart with each product, how it is made, its cost per liter, and its market size.
    • AIs work best if you go through the logic of what you want step-by-step.
    • It can take practice, but this approach allows you to “teach” Bing by asking it learn about topics, and then show you its progress as it works.
    • I can have it walk me through a simulated design thinking session (see below), or compare translations of poetry, or design new products, or do a SWOT analysis, and so much more.
    • The Bing chatbot, formerly known to some as Sydney
    • Because the process includes randomness, you may need to reset the chat several times (using the little broom icon) to get to a place where the system will work with you. You might also need to rephrase your requests. It is less likely to reject write a sample of a paper or write an imaginary draft paper than write a paper. You will need to experiment.
    • Sometimes I will experiment with Bing in one session, write down the prompts that work, and then do a new session with the recorded prompts to better use my six attempts. Bing also lets you pick how “creative” the answers you get are. After experimentation, I would suggest using Balanced when you want to work with numerical data (though Bing is still bad at math and still hallucinates) and Creative for everything else. Precise yields disappointing results.
    • Overall, Bing is immensely more powerful than ChatGPT, but also a lot weirder to use.
  • tags: chatGPT AI tools

  • tags: chatGPT AI Linkedin tools

  • tags: chatGPT AI Linkedin

  • tags: chatGPT AI Linkedin

  • tags: chatGPT AI Linkedin

    • In my position at NYU Stern, I think a lot about the future of education. 
      DO THIS: Tell ChatGPT to act as if it was giving a TED talk one year after the release of ChatGPT4. I asked the talk to be about the future of the MBA. ChatGPT talked about real time case studies, optimized research, and so on. I asked for 5 examples of one prediction around “personalized curriculum.” It detailed AI-driven skill assessment, adaptive learning modules, etc. Amazing.
    • Having trouble hiring the right person? 
      DO THIS: Detail to ChatGPT the exact responsibilities of the person you need. Stop guessing what characteristics you should interview for, or what questions to ask, or what job simulation you should give them. ChatGPT will give you a detailed list of qualities to look for.
  • tags: chatGPT Moodle AI

  • tags: digital badges taxonomy competency

Posted from Diigo. The rest of my favorite links are here.

No comments: