AI is not so much a tool that everyone uses in more or less the same way, but a mirror in which we see our own reflection--if we care to look.

Charles with an insightful take as usual.

The Model Context Protocol (MCP) promises to standardize AI-tool interactions as the “USB-C for AI.” While its simplicity accelerates adoption, MCP systematically overlooks four decades of hard-won lessons from distributed systems. This isn’t an academic concern: enterprises deploying MCP today are building on foundations that lack fundamental capabilities that every production remote-procedure calling (RPC) system, since 1982, has deemed essential.

The final word in AI haterade. This is an economic bubble, where NVidia sells GPUs so a handful of other companies can use them to lose massive amounts of money. It makes up a significant fraction of the economy, and it won't last forever.

The Unreliability of LLMs & What Lies Ahead

The only reason that LLMs took root in the first place was because our societies in the anglosphere have already developed cultures solely devoted to gaining status and keeping up the appearance of doing things rather than actually doing them. All other values, increasingly including even the accumulation of wealth (while this is still very much a thing that people pursue, wealth is increasingly becoming a proxy for status more than something desired in itself) are becoming subordinated to symbolic status games completely detached from anything real.

The word delusion keeps coming up when people talk about LLMs. Interesting.