I love my cousin to death, but she has a problem…
She'll ask you a question, one that ends with her giving you the answer she wants to hear. Then I respond in a nuanced way.
Maybe I disagree with everything she just said, or maybe I partially agree and think parts of her question miss the mark.
You know what she hears? That I agree with everything she just said, and she'll quote me as backing her up in her next argument.
Anyone who knows her well looks over at whoever she's quoting. That person usually shakes their head with a big NO, typically mouthing it at the same time.
You might see something similar happening in AI.
Unless you live under a rock, you've seen the big fight. One side says development needs to slow down, and the other says we have to stay ahead of China.
Both sides have a point.
Researchers just used Anthropic's Claude to break into OpenAI. AI 'jail breaks' keep popping up, notably the Hugging Face incident in July, when OpenAI's own models escaped a test and hacked another company.
Then you've got the AI agents that make us mad, the rising power bills and power usage, and don't forget the data centers.
Speaking of data centers, OpenAI caught China-linked accounts using ChatGPT to pose as Americans on social media and turn the public against them. US labs also accuse China's 'open models' of constantly mining our own AI to get better.
Now the argument has gone public.
AI creators from OpenAI, Anthropic and xAI keep trading shots with the chip makers, notably Nvidia CEO Jensen Huang.
Each camp sees what it wants to, and comes up with very different answers.
My cousin does the same thing when she tells someone I love her new haircut (I called it an abomination of hair).
Everybody has their own motivations. One camp wants more spending, another wants less, and a third wants to beat China.
Frankly, they're all right at the same time.
In a perfect world, all sides would get along. China would agree to a slowdown and we could all hold hands and sing Kumbaya together.
That's not going to happen.
It might take an actual national security scare to get all sides to find a solution. I hope it never comes to that.
I do know this much: AI keeps advancing for now, and I see plenty of money to make across the whole space.
I'll walk you through everything I see in AI this Wednesday.
Join me. [LINK]
Your only option,
Mark Sebastian