Never before has so much power been available to so many people, so inexpensively.
The difference between AI LLMs and PCs is ... AI can be used on a Smartphone. And BILLIONS of people have smartphones now.
Everything made from Silicon will get cheaper and more efficient over time. It's Moore's Law.
That's why I am not too worried about AI and data centers.
I think it's only a matter of time that AI solves the problems created by AI.
Remember when the media said nobody wanted to watch video on a phone because it was too expensive?
Now change that to “AI.”
It will look equally silly in 25 years time.
#AI isn’t an app, it’s not hardware or software, either.
It’s the “Intelligence Layer” for everything in modern computing.
Pretty soon, the phrase “#Artificialintelligence,” will go away and become as dated and redundant as “I” or “e” in Web 1.0.
“Sophisticated AI brand capture” is the new SEO. It’s curating.
You can’t just buy banners and AdWords anymore. Or stuff press releases and websites.
You need to generate more human interaction in your content, and the content needs real authority.
AI can crank out apps. That’s cheap. That’s easy.
But launching apps correctly, with GTM message, marketing strategy, early adopters, influencers, reviewers and press?
That’s expensive. And it still takes humans.
Some people immediately recognize the pattern because they’ve seen multiple platform transitions.
Only those who have actually experienced all 5 shifts will understand.
If you understand this meme that I created, let me know.
A humble message from a top #AI scientist, @JeffDean of @GoogleDeepMind to college grads:
“AI is an incubator for ideas, it cannot replace you and your human ingenuity.”
While the tools we use to build the future change, the fundamental driver of innovation is human creativity.
AI is an engine.
Humans must remain the steering wheel and the navigator.
Mapping that boundary is the act of navigation.
We are KNOWLEDGE NAVIGATORS in this new era.
“If you’re the smartest person in the room, you’re in the wrong room.”
Now with access to the frontier #AI models, you can be “in the room” with a higher intelligence.
@JeffDean@Google@NormJouppi@ieeemicro As AI models get bigger, they require staggering amounts of electricity. If Google keeps building data centers the old way, the power grid can’t handle it.
This paper highlights how they redesigned data centers and chips to make them vastly more efficient.