@TheAIShrink True but once you’re at that critical mass in terms of user facing feature - users will remain in ecosystem. It’s a race to build the best golden cage
Anthropic is setting itself up to be the next Google.
They haven’t been just a model provider for a while now. They’re fighting all-out to own the application layer: software, law, finance, scientific research, bio/medicine, and probably more to come.
Even though OpenAI was first to bring AI to the masses, Anthropic is the first to go all-in on the application layer. That’s where its success will come from.
And owning the application layer means owning the valuable data future models will build on.
Don’t be surprised if they’d announce a new chip or a deeper tie-up with Nvidia (investor) soon.
With a new model out every week, this song popped to my head.
“Slow down, you crazy child; And take the phone off the hook and disappear for a while; It's all right, you can afford to lose a day or two”
https://t.co/u0BsB1a5nC
Launched a new YouTube channel yesterday for my new podcast, Old School/ New Tech. If you're into cool tech from the perspective of someone who's been in the trenches for over 3 decades, you should subscribe (link in the first comment)
surprised more people aren't doing something like this
Codex now creates a "newspaper" for me every morning
Unread messages, calendar, surf report, news
Anything I can do to stay off my phone until later in the day is a priority
Secret detection as a learned layer, not a regex pass, is the right direction.
As agents get write-access to real systems, the guardrails have to be as smart as the agent.
This is the unglamorous infra that makes autonomous engineering shippable.
Glad someone's building it properly.