Silicon Valley remains a fascinating ecosystem where people with IQs of 160 and valuations of $60 billion regularly reinvent office politics, but with GPUs.
A few things that caught my eye recently:
1. The AI race has entered its "money has lost all meaning" phase
OpenAI reportedly raised around $110 billion in a private funding round at a valuation north of $700 billion.
What's remarkable isn't just the size. It's that investors are now treating frontier AI labs less like software companies and more like strategic national infrastructure. The conversation has shifted from:
"Can this company make money?"
to
"Can this company acquire enough electricity to continue existing?"
The hottest commodity in tech is no longer code. It's megawatts.
2. Google is openly redesigning the internet
At its recent developer conference, Google pushed heavily toward AI agents that perform tasks for users, monitor information, and generate answers directly.
This is potentially a bigger story than the latest chatbot benchmark.
For twenty years the web worked like this:
- User searches.
- Google sends traffic to websites.
- Websites make money.
The emerging model is:
- User asks AI.
- AI reads websites.
- AI answers directly.
Publishers are understandably staring into the middle distance and reconsidering their life choices.
3. Meta is throwing money at robots
Meta has been buying robotics-AI companies and investing heavily in humanoid systems.
The industry seems to have collectively decided:
"What if we put the chatbot in a body?"
A sentence that should probably have triggered at least one emergency meeting somewhere.
The long-term bet is that language models become the brains of general-purpose robots.
4. Startup drama is becoming performance art
One recent feud involved a founder accusing a rival startup of copying his company after the rival received much larger funding. Instead of quietly complaining, he posted a fiery video that attracted millions of views.
The lesson Silicon Valley appears to be learning is:
Attention may be a stronger moat than technology.
In the old days founders wanted patents.
Now they want followers.
5. AI companies are turning into consulting firms
Both Anthropic and OpenAI have been creating enterprise ventures and pursuing deployment businesses rather than simply selling models.
The realization seems to be:
Building the AI is hard.
Convincing a hospital, bank, or insurance company to actually use it is harder.
The boring implementation layer is where a lot of the money may ultimately end up.
My favorite underlying trend
The most interesting thing isn't a specific company.
It's that the tech industry increasingly resembles the railroad booms, electrification booms, and telecom booms of previous centuries.
Everyone talks about intelligence.
Everyone obsesses over models.
Meanwhile the real battle is becoming:
- Data centers
- Energy generation
- Chip supply
- Cooling systems
- Physical infrastructure
The future is arriving wrapped in science-fiction marketing, but underneath it is mostly transformers, power plants, construction crews, and accountants desperately trying to explain why the server farm requires its own zip code.
It's all very cyberpunk.
Except instead of neon-lit rebels hacking megacorporations, it's procurement managers negotiating transformer delivery schedules. A less cinematic species, but apparently just as important.
No comments:
Post a Comment