Infinite Loops Episode 166 - How AI Changes Everything Hosted by Jim O'Shaughnessy with guest speaker Ethan Mollick
[Professor Lazurus Reaction]
Artificial Intelligence has certainly changed an awful lot already. So much that I am writing this blog post reaction to Infinite Loops podcast Episode 166 immediately after watching it on YouTube so that it is full of flaws, run on sentences, and raw emotion. Otherwise I suspect there would be some people questioning if I wrote this myself, or asked an AI chatbot to do it for me instead. At this point, I can't even prove that I did it all by myself.
What a time to be alive.
Not long after ChatGPT started to get really popular, I spoke to an old college buddy who was teaching a course at Rutgers and we laughed as he explained to me that he was already catching students using the bots to write papers because they were being too lazy to even proofread the AI's work. The machine will often make obvious machine like mistakes, repeating the same word, or as they mention in this Infinite Loops episode, sometimes ChatGPT will just straight up lie to you. Replying to the users request with false and blatantly inaccurate information. At the same time, we both acknowledged how hard it would be to catch the student who does take just a little time to review the AI's work before turning the paper in. If the student takes the time to verify the information as accurate, are they still cheating on their homework? In the latter case, they still did the research, just in a non-traditional order of operations. And cheating on college papers, is just one tiny snow flake on the tip of the iceberg for the many early applications we've seen so far.
Though I have to admit, I noticed when I started posting AI generated images, a couple people I respected discontinued engaging with my online content. That hurt a little. But there was no way I'd let losing a few followers spoil my excitement while playing with my new visual arts toy. Finally I could come up with crazy crossover ideas, or cook up storyboard images and share them within minutes. That's when I realized the genie was completely out of the bottle and there was no stuffing it back in. Now I just patiently wait for the technology to develop drag and drop live action video to AI animation, so that I too can create feature length cinematic films, but without needing 100 million dollars for a production budget..
Ethan Mollick and Jim O'Shaughnessy discussed the incredible speed that achievements have been made within the AI space over the past year or so in the podcast, but what really got my attention was the chapter about "Using AI to explore liminal spaces" which starts around the 42 minute mark. Jim talked about a project that will feed every written text into an AI model and ask it to tell us about ourselves. I'd also propose a project that feeds a training model all known scientific and mathematical texts, then ask it to explain very specific details about the Universe. Then feed it all known religious texts and ask it to tell us about God. Then cross reference those answers and vectors to analyze the wealth of unbiased results they produce. Right or wrong, being able to juxtapose information in this manner could facilitate new scientific discoveries or even a spiritual revolution. The collaborative possibilities are endless. And don't get me started on the kind of immersive story telling and world building we could experience in a digital landscape created with a variety of user accessible AI tools.
There might be a handful of people who have read this far and are asking themselves, "What does this have to do with the Matrix?" Well the Matrix franchise teaches some valuable lessons for those willing to receive them. One of those lessons is a strong message about the power of teamwork. In both the Matrix and Terminator Universes we see examples of both good and bad Artificial Intelligence, working both for and against humanity. So I encourage people to work with the machine instead of against it. Take advantage of the cheap, and sometimes free opportunities to learn how these new tools can do work for YOU. At this stage of the game especially, while we know the machine can't think for itself, it is very important to make sure there are more positive use cases for AI than negative ones.
"The only way to prepare for the future is to embrace it." - Jim O'Shaughnessy
[Professor Lazurus Reaction]
Artificial Intelligence has certainly changed an awful lot already. So much that I am writing this blog post reaction to Infinite Loops podcast Episode 166 immediately after watching it on YouTube so that it is full of flaws, run on sentences, and raw emotion. Otherwise I suspect there would be some people questioning if I wrote this myself, or asked an AI chatbot to do it for me instead. At this point, I can't even prove that I did it all by myself.
What a time to be alive.
Not long after ChatGPT started to get really popular, I spoke to an old college buddy who was teaching a course at Rutgers and we laughed as he explained to me that he was already catching students using the bots to write papers because they were being too lazy to even proofread the AI's work. The machine will often make obvious machine like mistakes, repeating the same word, or as they mention in this Infinite Loops episode, sometimes ChatGPT will just straight up lie to you. Replying to the users request with false and blatantly inaccurate information. At the same time, we both acknowledged how hard it would be to catch the student who does take just a little time to review the AI's work before turning the paper in. If the student takes the time to verify the information as accurate, are they still cheating on their homework? In the latter case, they still did the research, just in a non-traditional order of operations. And cheating on college papers, is just one tiny snow flake on the tip of the iceberg for the many early applications we've seen so far.
Though I have to admit, I noticed when I started posting AI generated images, a couple people I respected discontinued engaging with my online content. That hurt a little. But there was no way I'd let losing a few followers spoil my excitement while playing with my new visual arts toy. Finally I could come up with crazy crossover ideas, or cook up storyboard images and share them within minutes. That's when I realized the genie was completely out of the bottle and there was no stuffing it back in. Now I just patiently wait for the technology to develop drag and drop live action video to AI animation, so that I too can create feature length cinematic films, but without needing 100 million dollars for a production budget..
Ethan Mollick and Jim O'Shaughnessy discussed the incredible speed that achievements have been made within the AI space over the past year or so in the podcast, but what really got my attention was the chapter about "Using AI to explore liminal spaces" which starts around the 42 minute mark. Jim talked about a project that will feed every written text into an AI model and ask it to tell us about ourselves. I'd also propose a project that feeds a training model all known scientific and mathematical texts, then ask it to explain very specific details about the Universe. Then feed it all known religious texts and ask it to tell us about God. Then cross reference those answers and vectors to analyze the wealth of unbiased results they produce. Right or wrong, being able to juxtapose information in this manner could facilitate new scientific discoveries or even a spiritual revolution. The collaborative possibilities are endless. And don't get me started on the kind of immersive story telling and world building we could experience in a digital landscape created with a variety of user accessible AI tools.
There might be a handful of people who have read this far and are asking themselves, "What does this have to do with the Matrix?" Well the Matrix franchise teaches some valuable lessons for those willing to receive them. One of those lessons is a strong message about the power of teamwork. In both the Matrix and Terminator Universes we see examples of both good and bad Artificial Intelligence, working both for and against humanity. So I encourage people to work with the machine instead of against it. Take advantage of the cheap, and sometimes free opportunities to learn how these new tools can do work for YOU. At this stage of the game especially, while we know the machine can't think for itself, it is very important to make sure there are more positive use cases for AI than negative ones.
"The only way to prepare for the future is to embrace it." - Jim O'Shaughnessy
NeoMatrixology
