it sure feels like we’ve reached peak AI. VC thought pieces and tweetstorms have hit all-time highs. crypto tourists have flocked to greener pastures. MBAs may soon outnumber nerds. is AI the new crypto? luttig.substack.com/p/is-ai-the-ne…

AI hyperbole in the tech industry is off the charts: foundation model FUD, incumbent invincibility, doomer declarations. most people have a story to sell you – or themselves – whether they admit it or not. I wrote about these hallucinations on a blogging site (link in bio):

Hallucinations in AI everything you think you know about AI has been told to you by someone with an incentive to lie (and mostly to themselves) by John Luttig Since the 1950s, AI has been a story of the boy who cried wolf — prophecies were over-promised and under-delivered.

hyperbolic narratives distort the AI discourse. it's hard to find signal in the noise. the real AI hallucinations are human: most people have a story of hope, cope, or mope to sell you – or themselves. I wrote about it in my first Pirate Wires piece piratewires.com/p/hallucinatio…

our chaotic AI discourse is entirely determined by incentives. do our expert talking heads believe they’ll be making billions from the AI utopia, or do they think they missed the opportunity? great piece on the powerful interests shaping our conversation from John Luttig

In 2023, Nvidia became the AI kingmaker. Thousands of companies and countries begged to purchase their GPUs. Will the insatiable demand persist? I wrote a piece exploring the Nvidia run-up, the resulting shortage, and the future of GPU demand blog.johnluttig.com/p/nvidia-envy-…

Devin is the first agentic product I’ve seen that successfully tackles multi-hundred step tasks by balancing long-term planning, skills, and memory. one of the few AI startups worth paying close attention to. congrats to Scott Wu and team, and grateful to have FF involved

Trae will never take credit for himself, but Newcomer’s profile on Trae Stephens is the first I’ve seen that properly acknowledges his role in founding the modern defense tech wave

despite recent progress and endless cheerleading, open-source AI is a worsening investment for model builders, an inferior option for developers and consumers, and a national security risk. I wrote about the closed-source future of foundation models here blog.johnluttig.com/p/the-future-o…

My favorite thing about John Luttig is that he’s always asking the hard questions, was fun jamming with him on this!

a16z simultaneously argues 1) the US must prevent China from dominating AI, and 2) open source models should proliferate freely across borders (to China) what does this mean? who knows. i'm just glad at Founders Fund we don't have to promote every current thing at once

Thoughtful contrarian take from John Luttig

This needed saying. The copium around open source AI is bizarre. Training costs are increasing by orders of magnitude. How many more do you think will be free?

turns out mark zuckerberg is not running a charity