AI round-up: Week of September 23, 2024

We will just jump right into the biggest news of the week, in my opinion: 
Sam (Altman) speaks.

First, here is some context worth your time, courtesy of Paul Roetzer.

Ok, now you can read Sam’s take on our future. (AI’s for everyone. On me!)

The big 5

1.    OpenAI is a weird company. 
Speaking of Sam Altman … when you stop and consider everything that’s happened over there in the past year, well, it doesn’t make you feel great about the stability of a company that seems to be shaping the future of society. (Casey Newton/Platformer)

2.    Apple Intelligence. Built on Google.
Well, this one seems a little weird. Here’s everything we—and by we, I mean Digital Trends—know about Apple Intelligence. At the end of the day, it sounds like my iPhone is going to run slower, listen to me more, record my data and still disappoint me with a new and improved Siri.

Other than that—and the fact that it’s built on Google’s Tensor chips—we don’t know much.

3.    AI will help decide if unemployed workers receive benefits.
I have a feeling we’re going to start seeing a lot of stories like this: AI is doing something, and a human will need to oversee what’s being done in the hopes of maintaining accuracy and fairness.

If you’ve heard me talk recently, you know I follow the Ethan Mollick school of thought and focus on the ‘human in the loop.’ See if you can spot that coming to life in this read.

4.    80% of brands are worried about agency use of generative AI.
They should be. And here’s why. The real issue is of the companies surveyed in this poll, only 38% of them have shared AI usage terms with their agency partners. And only 29% of them have reviewed all agency contracts to identify and add AI-specific clauses.

5.    YouTube can now generate six-second videos.
In my opinion, here’s the big takeaway with all of these platforms/tools: we are seeing the future of shows/entertainment materialize right in front of our eyes. If I have an idea for a character and want to develop a short series, I can do it now. I don’t need it to be picked up by a network, and I can enter into whatever advertising or sponsorship agreements I want if enough of you like it. We are seeing a new era for the creators.

Question though: if something goes up … what comes down? Networks? Studios? Actors? All of them?

Learn a little

In the world of AI, there is a thin line between ‘learning’ and ‘freaking out.’ This is one of those.

ChatGPT’s advanced voice feature is finally rolling out to Plus and Teams subscribers. Cool, right? We can finally talk to AI as if they are sitting there next to us—working with us. Not weird at all.

Or we can do mind-blowing things like this (courtesy of Ethan Mollick).

Did you hear about…

…Audible is going to start using AI for select audiobooks. (Bloomberg)

…A recent study of large LLMs shows that they will give wrong answers rather than admit they don’t know. (Nature).

…8 companies and the U.S. government have formed the Partnership for Global Inclusivity on AI. (U.S. Department of State). I don’t really know what this means.

…OpenAI’s communication officer leaving the company? (The Verge)

…OpenAI is looking to abandon its non-profit status and give Sam Altman equity? (Reuters)

…AI can pass Google’s CAPTCHA test. (Decrypt) Great. One of the few things I was holding onto—that I’m really good at finding all the traffic lights in a picture—is one more thing AI can do.

…Doctors using AI to write follow-up messages and reports for patients? (New York Times) At least you’ll be able to read it?

Must read/must discuss:

“If journalism is going up in smoke, I might as well get high on the fumes.”

So says this anonymous author in their article for The Guardian. Why? Because they provide a human response to simulated prompts that are then used to train AI LLMs.

This article is a little too much and I’m not sure how I feel about it. Can you stop the inevitable? More importantly, do you want to be part of the solution or the problem? Hard to tell when you can’t tell which is which.

Did you miss MAICON?

Don’t worry! We have you covered. On October 8, from 4:00 p.m. to 6:00 p.m., we’ll be meeting in person (as in ‘the same room’) to discuss MAICON takeaways. I’ll share some takes and open the floor to others who attended. Don’t miss the chance to talk AI, hang out and ponder our futures!

-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.