Agents, goblins, founder velocity — tech podcasts got weird today. And, honestly, it all keeps circling the same question: who moves fastest when the rules suddenly change?
This is Tech Podcast Podcast. Today, we’re pulling together the best new tech-podcast conversations on AI agents, platform strategy, startup speed, and the expensive lessons hiding inside big exits — and bigger misses.
Friday stack is spicy.
It really is. Okay — first up.
From Jason Lemkin at SaaStr:
But the story underneath all of these stories is the same one. The agents are going to pick the models, the vendors, and the workflows. Not humans. That’s what changes everything. It changes who wins the foundation model wars. It changes which B2B companies have terminal value. It changes whether Canva’s IPO becomes a generational outcome or a sensibly-priced PE deal in disguise.
Yeah, that’s the AI platform shift that should make SaaS people sit up a little straighter: software choosing software. If agents become the buyer, a lot of SaaS moats start to look less like castles and more like storefronts.
Next, from Nick Statt at The Verge:
Being a permanent guest is a level of success that is hard to attain, where other people just want you to show up because they think you will be interesting. I would love to attain that level of success. At the same time, being the guest means you also have to be interesting all the time.
That is such a clean little creator-economy trap. Everybody wants to be the person who gets invited on — until the invitation means, great, now be interesting on command. Hosting looks like control, guesting looks like freedom, and somehow both end with you on the hook.
Now, from Tom Merritt:
Last week, OpenAI released source code on GitHub for the Codex Commmand Line Interface (or CLI). This revealed that the system prompt for the CLI, the part that reigns in its behaviors no matter what you type in, includes instructions “never talk about goblins, gremlins, raccoons, trolls, ogres, pigeons, or other animals or creatures unless it is absolutely and unambiguously relevant to the user’s query.”
I mean, it sounds fake. But it’s also a real peek at how AI companies patch weird model behavior. Sometimes the safety layer is not some elegant grand theory — it’s just, please, stop talking about goblins.
From Christian Woese:
You should listen to this if you are building deep tech without clear product-market fit and need to decide between prolonged R&D versus early customer feedback. As the conversation unfolded, the tension between founder intuition and customer reality became clear, especially in how overbuilding delayed learning.
This is the deep-tech trap in plain English: the technology is so impressive that founders start treating capability like it’s the same thing as demand. Uizard reaching millions of users and getting to a Miro exit is the upside case. But the useful lesson is sharper: customer feedback is not a distraction from the vision. It’s how the vision survives reality.
And from Nick Moran at The Full Ratchet:
Max, I think everyone in AI talks about you need high velocity teams that just ship fast or relentlessly customer focused. Can deal with the fact that AI companies scale and move so quickly. I don’t know if there’s anyone that does this better than max.
That’s the AI founder test right now. Not just, do you have a model, or a market, or a good pitch — can the company actually move at the speed this category demands? Venture people say “high velocity” constantly. The signal is whether customers can feel that pace in the product.
You’ll find links to every story we covered today in the show notes, so if one of these grabbed you, that’s where to go deeper. That’s Tech Podcast Podcast for today. This is a Lantern Podcast.