Inside the AI "Frankenact" Disaster & The Fight for Developers w/ Jake Ward
TL;DR
Policy is still chasing a moving target — Jake Ward says the core fights since 2012 have barely changed—privacy/data and competition—but AI has accelerated everything while lawmakers still misunderstand how developers actually build, ship, and scale software.
Europe’s AI and privacy rules created a ‘Frankenact’ problem — In Ward’s telling, GDPR kneecapped Europe’s adtech startup scene and the EU AI Act got warped by politics and generative AI arriving mid-process, spawning a compliance cottage industry instead of more innovation.
State-by-state AI regulation is a bad fit for software speed — Using Utah and Colorado as examples, Ward argues early fragmented AI rules are ‘terrifying’ and calls for a federal framework focused on consumer harm, discrimination, downchain liability, and protections for open-source and model builders.
AI won’t kill engineering—it will squeeze juniors and raise the value of veterans — Ward says senior engineers are more valuable than ever because you still can’t ‘vibe code’ core engineering decisions, even if AI is already blowing junior prototyping and boilerplate work ‘off the table.’
The scarce skill is no longer just coding, but taste and communication — Both Ward and Reis keep coming back to discernment: knowing why something feels right or wrong, being able to explain it, and resisting the temptation to let Claude or ChatGPT do all the thinking for you.
The next big test is what consumers actually want from AI — Ward is watching whether people accept ads in AI products, pay $20 for better models, embrace Meta’s Ray-Bans as ‘ocular superpowers,’ or split into two camps: people who want bots for everything and people who still want humans.
The Breakdown
Why the Developers Alliance exists at all
Jake Ward opens by tracing the Application Developers Alliance back to 2012, when he and his co-founder realized policymakers had almost no idea who app developers were or how they worked. The group grew to 160 corporate members and 150,000 individual developers, all aimed at getting builders a seat at the policy table instead of letting regulations get written by people who, as Ward puts it, “just don’t know what they’re talking about.”
The two policy fights that never went away: privacy and competition
Ward says the big themes have stayed oddly stable for more than a decade: data/privacy on one side, competition on the other. He points out the U.S. still has no federal data privacy law—just a patchwork of roughly 29 state laws—while Europe’s GDPR had a “profound effect” on innovation, access to capital, and startup exits, especially in adtech.
How Europe ended up with a regulatory ‘Frankenact’
Joe Reis connects this to his conversations with EU AI Act architect Kai Zenner, describing how a once-clean proposal got pulled apart by member-state politics, then blindsided by generative AI. The result, in their telling, is a muddled regime where consultants now help companies get “ready for regulation instead of innovation,” and Ward doesn’t buy that Europe’s recent simplification efforts amount to true deregulation.
Why Ward fears fragmented AI rules more than AI itself
When Utah’s AI bill comes up, Ward calls state-by-state frameworks dangerous because software moves too fast for Colorado and Utah to govern differently. His preferred answer is a federal baseline: ban discriminatory practices, clarify downchain liability, protect open source, and avoid crushing model developers with civil liability before the market even reveals what the technology can do.
The accountability problem: from Waymo crashes to algorithmic discrimination
Ward gives a clean explanation for why lawmakers keep reaching for AI-specific rules: if accountability gets blurred, responsibility feels impossible to assign. He uses Waymo as the easy example—if a self-driving car hits someone, who’s actually at fault?—then extends the logic to credit, surveillance pricing, and other systems that can smuggle prohibited discrimination into decision-making at machine speed.
AI coding tools are changing who matters on engineering teams
This is where the conversation gets more personal for developers. Ward says AI is “the greatest tool we’ve ever had,” but also bluntly argues that junior engineers are getting displaced first, while senior engineers—the people who know which button to push and why—are becoming even more important because product-quality engineering still can’t be automated away by vibes.
Developers probably won’t unionize—and that says a lot about developer culture
Asked whether AI anxiety could push developers toward unions, Ward laughs at the irony: people asked in 2012 whether the Alliance was secretly a union. His answer is basically no—developers are too individualistic, too freelance-friendly, and too wired to build things alone, then argue online about how everyone else built them wrong.
Taste, communication, and the human-speed skills AI can’t replace
The back end of the talk shifts from policy to craft. Ward says the most durable skills are communication and “taste,” which he frames as the courage to have an opinion and feed it; AI can draft, summarize, and brainstorm, but it can’t supply judgment if you’ve got nothing in your head to begin with.
The next real battle: what kind of AI experience people actually choose
Ward closes on consumer behavior, not model benchmarks. He’s watching whether users accept ads inside ChatGPT, pay for premium assistants, or embrace devices like Meta’s Ray-Bans as everyday “superpowers”—because the market is about to split between people who want bots for everything and people who still get mad they can’t order at McDonald’s from a human.