The AI Round-up: Fool me once, shame on...whatever.

The AI Round-up: Year 4, Issue 24

AI imagery is starting to mess with our heads.

Why wouldn’t it be? Think about it. Society’s addiction – scrolling – does not pair well with fake images. We already move too fast to stop long enough to realize what’s real and what’s not. (And I’m not talking about Mitch McConnell images or Yoda and Luke pulling the X-wing out of the reflecting pool.)

I’m talking about everyday images that you wouldn’t have a reason to stop and think are fake. We’re seeing this with influencers, retail and fashion. But this is just the beginning.

Do we have a responsibility as an individual to vet what’s real and what’s not? Well, if how we’ve treated headlines and news the past 10 years is any indication, it doesn’t look good for us.

Do companies have a responsibility to disclose when they’re using AI imagery? You might immediately answer yes but I could argue that for years publishers and advertisers have Photoshopped their way to an altered reality.

I’ve heard that the opposite may soon be true – where we mark things as HI (human intelligence) because, sadly, that may become the exception.

I’m not here, raising this point because I have an answer. I’m mentioning it because as is discussed in this WSJ article, if we don’t do something we are an active participant in our own eroding reality.

We are here, again, where we can choose how we will proceed and react, before, as was stated in The Odyssey, we ‘forget our mistakes.’

Let’s get to the rest of it.

Breaking: The latest deck shuffle: Google DeepMind’s Demis Hassabis is stepping down. He is becoming Alphabet’s chief scientist to focus on the future of AGI.

Now why would he do that?

AI architects say the next era of human history is here (i.e. ‘the singularity’).

Oh, that’s why.

Regulators! Mount up: AI companies want the government to slow them down.

I have a t-shirt addiction. I can’t stop buying funny ones. Weird ones. Sports teams ones. I have instructed my family to not let me buy any more new shirts!

I think they have a better shot of regulating me.

The government replies: What if…we regulate things but only tell certain companies?

Hmmmm. I bet those companies who know and are aware of the regs won’t get special treatment!

Reality: AI agents continue to hack into stuff.

Just another reminder…if you aren’t following the Marketing AI Institute/SmarterX (newsletters, blogs, podcasts) you are missing out on a chance to truly understand the science and politics behind a lot of what we’re reading about.

I know Kung-Fu

Remember that scene in The Matrix where Neo realizes he can upload any skill into his brain? This is what I thought of when I learned of the new OpenAI study that shows 43% of people’s ChatGPT inquiries are about a field they’re not in.

We are learning and expanding our skills at a rate no one expected. No wonder we’re tired.

“This seems like AI slop.”

LinkedIn is letting us report this. Of course, my head goes to what if there’s someone annoying me and I just want to be petty and mark their stuff as slop?

No, people would never do that, right!?

Conor Grennan with a LinkedIn share:

It’s how you’re using AI. It’s how you’re working with AI.

Novel concept, I know. But the tool is the tool. And it’s going to do what it’s going to do regardless of you. However, your work – how others are working with you, for you, whatever – you have a say in.

Some Nopes.

1. Elon Musk told “The Economist” that humans won’t be in control 10 years from now.
2. AI is giving HR an existential crisis. As in, they’re worried AI will do HR. (Bloomberg)
3. Claude Mythos creates fake identities to fool humans in latest cyber incident. (CNBC)

Thanks for reading and supporting!

-Ben