AI round-up: Week of Nov. 27, 2023
“I am not sure who said it first, but there are only two ways to react to exponential change: too early or too late.”
--From Ethan Mollick’s blog, One Useful Thing
So much to untangle from the last week. Let’s see if we can make sense of everything that’s happened while sharing the hottest stories from the past week.
This week’s issue: 14 stories
The AI anxiety meter: High
How long to read: A good 30 minutes…not counting podcasts.
To note: My role in this is serving as an aggregator. Reading, listening, absorbing and trying to piece all of this together in a way that keeps you plugged in and up on what’s most relevant. It is becoming increasingly difficult to do so without adding some editorial, but I will do my best to maintain the objectivity we established early on.
There and back again. No, we’re not telling a story about Samwise Gamgee. We are telling the story of Sam Altman. Last we spoke, his ousting as CEO of OpenAI was your Thanksgiving dinner appetizer. But a lot has happened since then. Let’s recap:
1. If you didn’t get to listen, Sam Altman was removed as OpenAI’s CEO, and there is no better breakdown/coverage of this than Paul Roetzer’s solocast on the Marketing AI Show pod. (Episode 73)
2. At that same link, you can listen to this week’s podcast where Paul talks about Sam’s return. (Episode 74)
3. The details of his return are still somewhat unknown but Sam did give an interview to The Verge (although he didn’t say much). This is the first – and only – interview so far.
4. This is a bit in retrospect, but The Hard Fork podcast did feature an incredible interview with Sam that they aired as part of their take on what happened when he was dismissed.
So, what does it all mean? Well, if you read, listen and try to connect some dots between the lines, it seems that … something happened. At some point this year. That was severe enough to rattle some cages and spur the influencers I follow to put forward some interesting theories.
For us, I would say it means regardless of your position, planning for where we are and where we ultimately are going, is a top priority. (We can discuss what that looks like, but not in this newsletter.)
So, where are we going and when will we get there?
5. Nvidia CEO Jensen Huang says AI will be fairly competitive with humans in five years. It’s worth the quick read as to why he thinks this. And it's worth stepping back and asking yourself, “Is he saying five years to avoid alarm, or does he really believe that?” (In light of everything you read about why OpenAI temporarily broke with Sam Altman.)
6. I know I’m hitting you with a lot of listening, but you can always come back to these … here’s a recent podcast on HBR’s IdeaCast where they talk to Jensen about his take on running an AI-led company.
7. Speaking of an AI-led company, did you know no one has truly figured it out yet? Which is why Ethan Mollick’s blog “Reshaping the Tree: Rebuilding Organizations for AI” is such a compelling read.
8. We’ve made it to story eight and have yet to mention that today is the one-year anniversary of ChatGPT. That’s how thick the AI world is right now. But it is one year old today, which is just so wild to think about how fast and far things have moved in the past 365. Christopher Penn talks about this very thing in his blog "ChatGPT turns 1. What have we learned?" We’ll make it this week’s must-read.
9. On a side note, I did ask Claude to write a birthday toast to ChatGPT.
Me: It's a birthday party. Your good friend ChatGPT is being celebrated for turning one year old. The guests start clinking their glasses and ask you, Claude, to make a toast to them. What would you say?
Claude: raises glass Dear friends, we are gathered here today to celebrate a very special birthday for our friend ChatGPT. Though only one year old in age, his vast knowledge and helpful nature make him seem much wiser.
ChatGPT, we thank you for always being there when we have a question that needs answering or just want some friendly conversation. Your talents continue to impress us daily. Here's to many more years of learning and growing together as friends! Happy birthday! smiles and takes a sip from glass
10. SI and AI. This week’s “What is this, now?” article is about Sports Illustrated. They thought it would be a good idea to use AI to create a reporter, a backstory and have that reporter publish articles. Oops.
11. Did SI have an AI policy? Probably not. But as it turns out, having one doesn’t always matter. A recent survey revealed more than half of employees are using unapproved AI tools in some form anyway.
12. I know most of us reading this are thinking of that last story from a comms standpoint – as in, what are people writing and creating using these unapproved tools? And that’s worth worrying about. But what about other functions within a company using these unapproved tools … or skirting compliance regs to do it? CSO Magazine looked into this and found that nearly 100% of coders and developers are using AI tools – and 80% are bypassing security policies to do so.
13. Whoa. This article from MIT Technology Review on how Google’s DeepMind helped create more than 700 new materials is above my pay grade. How do you even wrap your head around this one?
I know I hit you with a lot. So, let’s end here:
14. The word of the year is authentic. Hey, at the end of the day…you be you. No AI can take that away from you. Well, aside from Deepfake and AI cloning. But you get the point.
As a reminder, this is a round-up of the biggest stories, often hitting multiple newsletters I receive/review. The sources are many…which I’m happy to read on your behalf. Let me know if there’s one in particular you’d like me to track or have questions about a topic you’re not seeing here.