Collaborative Intelligence Week in Review - 18Aug2023
Best business article(s) I read this week…
Ideas are Dimes a Dozen: LLMs for Idea Generation in Innovation
This working paper from Wharton presents some of the most intriguing research I have read this year. It provides a great example of Collaborative Intelligence in action: pairing humans and intelligent machines to improve performance in pursuit of a human-specified goal. (Full Disclosure: I have taken several innovation courses from Karl Ulrich, one of the authors.)
The authors created a competition for generating innovative ideas. They pitted students in a well-regarded innovation course against ChatGPT-4. Here’s what they found: “ChatGPT-4 can generate ideas much faster and cheaper than students, the ideas are on average of higher quality (as measured by purchase-intent surveys) and exhibit higher variance in quality. More important, the vast majority of the best ideas in the pooled sample are generated by ChatGPT and not by the students.”I included an article on a similar topic a few weeks ago. This is the cover story for the most recent Harvard Business Review. Re-skilling is gaining more C-Suite attention and generating intense competition for AI talent.
Best technical article(s) I read this week…
AI for the Generation and Testing of Ideas
This isn’t a particularly technical article. The authors propose an AI-based Knowledge Development Environment (KDE) for generating and testing new ideas. Their Generate And Search Test (GAST) algorithm allows individual knowledge workers to “efficiently create solutions that here-to-fore [sic] required the best collaborations of the best experts.”
Other item(s) of note…
Generative AI Hacking Challenge Finds Security Vulnerabilities in LLMs
If you’re an executive trying to understand the benefits and risks of Generative AI, this article is worth reading. Ask your CISO to explain the technical details.
The coolest thing I saw…
Neuroscientists Recreate Pink Floyd Song from Listeners’ Brain Activity
I may be the only one who thinks this is cool. I have long had an interest in neuroscience. AI is advancing many areas of neuroscience research.
Here’s a bonus entry in case you don’t care about neuroscience. Meta recently released three AudioCraft components. The first generates music from a text prompt. The second generates other types of sound from a text prompt. The third is a new Codec for formatting AI-generated audio. Click on the various page links to see demos.