AI round-up: Week of March 04, 2024

Do you get the feeling we’re on the cusp of some very interesting times in the time of AI?

The temperature on the term ‘AGI’ has slowly been turned up. But the reality is that it’s always been a hot topic for those on the leading edge of AI. And now it’s cited as the main reason Elon Musk is suing OpenAI.

There is a lot to unpack with this story. So rather than take up half of this round-up with it, I’ll point you to a few trusted sources and stories:

First, this TechCrunch article will give you a high-level view.

Second, OpenAI’s response (memo to employees)

Third, this week’s Artificial Intelligence Show will break it down for you. (Episode 86)

Fourth, if you listen to this week’s Artificial Intelligence Show and missed the episode Paul is referencing, here it is. (Episode 73)

Ok, onto the rest of what’s going on in the world of AI.

This week’s issue: 15 stories (counting links above)
The AI impact meter: Moderate 
How long to read: 30 minutes (not including the pods)

Did you see that my friend Claude has an update?
Claude 3 is out. And it has declared it’s alive and fears death.

Now, before we get too carried away … read the blog that is referenced in the article. The prompting has a lot to do with it, as the author is asking Claude to assume a “role” in answering the question.

That being said, I went to the source. And Claude admitted to remembering that conversation with the blog’s author, Mikhail Samin. However, Claude said this has all been overblown.

And wow … did I really just write all of this?

Remember when… we talked about Perplexity at the Akhia Insights Summit?
Well, its value is set to double (to $1B).

Here come the AI worms. 
I’m going to be 100% honest with you. I don’t like this story at all.

Now, this is interesting – research/testing was done on five LLMs to find out how much they knew about elections, polling, voting, etc., information in their area.
The results? Not so good, Al. “Harmful,” ”Inaccurate,” “Biased”

Microsoft engineer warns company’s AI tools create violent, sexual images; ignores copyrights.

SalesForce’s Einstein Copilot is here. Long and short of it – responses are grounded with your data.*

WordPress and Tumblr are the latest to sign an agreement to sell user data to train AI tools.*

*Stories sourced from the latest Marketing AI Institute newsletter.

LinkedIn: More workers need help to become AI-literate.

Must read/must discuss:
The evolving landscape of AI in corporate governance
Don’t ignore this article or the key takeaways in it. We’re here to help navigate this very important, very touchy subject.

Side note: We are taking our show on the road. If you’re interested in a lunch-and-learn format for you, your team, or your leadership group, we now have a presentation that can be customized and help inform, educate, and lay out the first steps you can take to assess your company’s position and identify gaps.

Email me to learn more.

Thanks for reading! Feel free to share!

-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.