Ed Zitron Unfiltered on OpenAI, Anthropic & Why the Whole Thing Is a Con
Categories: VC, Startup, Product
Summary
Ed Zitron argues the AI boom is fundamentally a financial con with no proven ROI—companies like OpenAI and Anthropic only discuss theoretical value, not real deployments, and the fact we're still debating whether AI 'works' proves it doesn't.
Key Takeaways
- The core con is financial, not technical. If AI had real ROI, companies wouldn't need to constantly defend its value with theoretical arguments—proven products don't require philosophical justification.
- AI-generated content is detectably lifeless and mediocre. Even boring human-written LinkedIn content has human emotional resonance, while Claude-generated text feels empty and follows predictable linguistic patterns.
- The debate itself is the tell. We shouldn't still be asking 'does AI work?' for something genuinely revolutionary—the persistent need to convince people signals the technology hasn't delivered tangible value.
- Distinguish between narrow technical capabilities and real-world reliability. LLMs can assist with code but are unreliable and unstable at scale—the gap between marketing and actual deployment is massive.
Related topics
Transcript Excerpt
Today on the podcast, I finally bring Ed [music] Zitron around on AI. >> Lifeless is empty. Look like a dog's [ __ ] in Thanksgiving. Stupid. You [ __ ] Why don't you like data centers? >> Okay, maybe not quite, but I do have a lively debate with Tech's loudest AI skeptic. >> So, how are you going to get around the um return on investment thing? He's like, "Yeah, somebody will work it out." Sammy Clammy Sammy, this is your job. From his podcast, his newsletter, and his 18,000word takedowns, he has made one thing very clear. AI is a bubble that's about to burst. I'm Eric Newcomer, author of the newcomer Substack. Let's get into it. [music] [music] >> Ed Zitron, welcome back to the Newcomer Podcast, round two. >> Thanks for having me. >> We're live in person, just you and me, mono. the uh I …
More from Newcomer
- NVIDIA Needs to Sell $250 Billion in GPUs Per Quarter. That's Impossible.
- Amanda Askell on AI Consciousness, Claude & Silicon Valley’s Biggest Fear
- Tony Fadell Unfiltered on Apple, OpenAI & the Next Big Device