AI round-up: Week of August 5, 2024
I always feel bad when I’m a day behind in sending this out. But then I do that last-minute scan of the most recent AI news, discover a hot topic and instantly feel better that I can bring this pressing news to you in real time.
Now, while this first topic isn’t tied to a new tool or a huge shake-up, it does speak to one of the underlying concerns a lot of folks have had around the growing use of AI: addiction.
Ethan Mollick talks a lot about resisting the anthropomorphism of AI – assigning human traits, feelings, and behaviors to inanimate objects, nonhuman animals or nature.
But dammit, if that isn’t hard to do. And honestly, the toothpaste may already be out of the tube on this one.
The Big 5
1. OpenAI warns users could become emotionally hooked on its voice mode.
There is a system card in OpenAI’s new GPT-4o that reveals some concerns around addiction to voice mode. This was revealed in the system card’s ‘Anthropomorphization and Emotional Reliance’ section. This piece is from Wired.
So be careful. And remember … AI isn’t real! (Yet.)
2. This warning makes me question if the OpenAI leadership team is growing increasingly concerned about the future of the company and its approach to AI. Nah. I’m sure the president taking a sabbatical and two other senior leaders leaving have nothing to do with anything. (Mashable)
3. Oh, this one stings a little bit. Christopher Penn wrote about the ROI of AI. And it’s terrific, as usual. But the biggest takeaway of his piece is directed squarely at us – the marketers – when he says, ‘You need to know what you were already spending before you can determine if you’re spending less.’
4. Will AI kill meaningless jobs? And is that so bad?
You know who might think it’s bad? The people who do those ‘meaningless’ jobs. (Side note: while I think there is a great benefit to eliminating or automating mundane tasks, I’m not loving the title of ‘meaningless.’ But hey, far be it for me to critique The New York Times.)
5. Colleges race to ready students for the AI workplace.
We’ve come a long way from telling students not to use A!, as you’ll see in this WSJ article. Someone better let OpenAI know since they’ve developed a tool that will reveal cheating students. Surprisingly they are hesitant to release it. Hmmmm.
Learn a little
I came across this in preparing for an AI workshop. I immediately saved it and knew I would be sharing it with you today:
The AI-powered content strategy: 13 prompts for more accurate targeting. (Courtesy of the wonderful Andy Crestodina and Orbit Media Studios)
Did you hear about…
…Nvidia caught scraping data from Netflix and YouTube again? (digital trends)
…JP Morgan giving all of its employees an AI assistant powered by ChatGPT? (CNBC) I can’t be the only one thinking this is a big experiment, right? Also, a hell of a way to give ChatGPT some terrific real-time training data!
…struggling AI companies looking to big tech to bail them out? (WSJ) Smells like a Monopoly cooking, to me.
…Boston partnering with Google AI to improve traffic patterns? (Quartz)
…AI’s ability to help Coca-Cola place a digital beverage next to any meal? (digital trends) As Ethan Mollick said, you can no longer trust your own eyes. My oh my. Make mine a Jack and Coke, please.
Must read/must discuss:
“There are things I’m very worried about for the future in AI, and then there are things I think are just amazing and unparalleled in the possibilities of what we’ve had before. So I think we need to live in both of those realities that there are fears we have to address and we have to think about regularly, but there are these amazing things that can open up possibilities, and we can’t lose sight of that, and we can’t become restricted by our fears and not take advantage of the opportunities ahead.” – Paul Roetzer
This is the opening of the latest Artificial Intelligence Show (episode 109), where Paul and Mike do a deep dive into their State of Marketing AI Report.
It’s a must listen.
Thanks for reading!
-Ben
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 you’d like me to track or have questions about a topic you’re not seeing here.