Collaborative Intelligence Week in Review - 28Jul2023
Best business article(s) I read this week…
AI Leapfrogging: How AI Will Transform “Lagging” Industries
Interesting article from seed-stage venture firm NfX that gives their take on where and how the best opportunities for AI will emerge.
If you are an active investor, you may be interested in this. The authors describe a process for using Large Language Models (LLMs) to examine the content of the verbal communication that happens during earnings calls to unveil hidden insights that can inform investment decisions.
Best technical article(s) I read this week…
Instruction Mining: High-Quality Instruction Data Selection for Large Language Models
The authors propose a method to select examples of natural language instructions that are used to fine tune Large Language Models (LLMs). They claim that this method improves results in 42.5% of cases vs. models fine-tuned on unfiltered datasets.
Other item(s) of note…
Why Generative AI Won’t Disrupt Books
I am an avid reader. If you are, too, you will be interested in this article. It provides a history of various technological developments and their impacts—real and imagined—on books.
Apple Moves into Generative AI with Its Own AJAX Framework
This is a short video from Bloomberg Television replaying their “scoop” about Apple’s plans for generative AI.
Technology trends that matter most in 2023
Last week, I included a similar topic from MIT Technology Review magazine. This list was constructed by McKinsey Digital. The link above takes you to the Executive Summary web page. I also recommend the full report. It can be accessed via the Download button on the page.
Depending on your view of AI, you may or may not think this comic is funny.
Coolest thing I saw…
AI-Generated “Barbenheimer” movie trailer
This movie trailer went viral last week. The background story is available in this VentureBeat article. I am not endorsing the specific trailer. I am endorsing the end-to-end production using AI tools and Adobe Creative Cloud.
Make sure you view the video from the beginning. When I run it, it starts at the 0:24 mark.