Two Bits

Two bits of information that I’m chewing on:

First, as I alluded to in my last send, the bridge (or, perhaps, outgrowth?) between knowable and unknowable models is firmly emerging as LLMs, and their accompanying natural language processing capabilities.

The defining characteristic of this LLM-software bridge is that English is rapidly becoming the new default programming language. This in no way implies that programming languages are going away (or that developers won’t need to know them). Rather, it means that with the advent of this new wave of software tools, English is the meta-language that is used to control (or write) other underlying languages.

To that end, this is an excellent essay on that development (and more), entitled “Software 3.0.” Of course, the title has me wondering if we won’t be at software 12.0 by next October.

Second, and I mention this only because it keeps gnawing at me, but has anyone else been a bit surprised by the lack of Apple announcements around AI? Especially as we’ve seen Google (with Bard and the Replit announcement), and now AWS (with Bedrock) begin to seek more prominence, Apple’s relative silence is starting to get my attention.

And that brought me back to something Andrej wrote. Namely, that the world will change as low-powered intelligence becomes pervasive via pre-trained neural networks running on small, inexpensive chips. So now I’m wondering if Apple is working on something akin to Andrej’s premonition. To be clear: I have no inside knowledge of anything, I’m just spitballing given the current state of affairs.

More soon…