34:58
AI and product strategy
In the current AI inflection point, building has stopped being the bottleneck and deciding what to build has become the bottleneck, so a product's real moat is customer-grounded value and accumulated market experience rather than the bare fact of 'having AI' — and alongside this shift, product people and organizations need to consciously protect the human, joyful side of work that AI hype and remote work are eroding.
Building is no longer the bottleneck in an AI world; deciding what to build now is.
Positioning has become harder because tools like Lovable and 'cloud/plug code' let anyone clone a product's surface in a weekend.
SaaS companies' durable differentiator is experience built across a broad market, not the presence of AI itself; 'AI is not a product in itself.'
AI is 'just another technology,' compared to prior hype cycles (blockchain, big data, apps) but expected to have more lasting, internet-level impact.
Conference conversation has shifted from 'AI kills our jobs' toward product and design mattering more when building is cheap and fast.
A parallel theme at the conference is preserving the human element of work — joy, fun, connection — under pressure from both AI hype and remote work.
Productivity gains tend to get reinvested into higher output expectations (build 2x or 3x more) rather than banked as personal time or creative space.
Conference-attendance advice: don't try to absorb every talk; pick one or two that resonate and extract two or three takeaways.
Advice for aspiring speakers: start at a local Product Tank meetup, build talks around personal stories, and don't imitate other speakers' styles — 'be yourself.'
In legacy, non-AI-native organizations, product people should stop 'teaching' product methodology and instead use the organization's own language and just start doing things.
Positioning fundamentals haven't changed with AI: products still need a clear reason to exist beyond 'we use AI.'
Predicts roughly 12-18 months more of AI experimentation ('throwing stuff at the wall') before things steady out; the traditional team ratio of about seven engineers to one PM to one designer is currently unsettled.
Eric Ries's new book 'Uncorruptible' argues standard corporate governance and IPO mechanics strip the 'special' quality out of once-loved companies, profiling alternative ownership structures (REI, Patagonia, John Lewis) that resist this.
The 'producty' framing itself — calling something a pilot, talking about iteration — triggered resistance from an operations team; reframing the identical ask as 'come look at our new shiny thing' got buy-in where product jargon had failed.
The anxiety that 'everyone can now build their own tools' is partly a tech-bubble illusion: most people (her 'mom' example) don't want to manage their own integrations or MCP servers, so packaged products retain demand regardless of do-it-yourself feasibility.
Rising day-to-day team conflict is attributed less to AI disruption and more to remote work removing the informal in-between-meeting moments that used to defuse friction before it accumulated.
Positioning a product as 'AI first' can be a liability rather than an asset outside the tech world, because many people conflate AI with decades-old automation they already dislike.
Eric Ries's book reframes conventional startup-success milestones (adopting standard governance, IPOing) as the very mechanism that extracts value and hollows out what made a company special in the first place.
Building several SaaS-replacement tools in-house doesn't just cost engineering time — it forfeits the cross-company pattern-recognition that dedicated SaaS vendors accumulate by serving many customers' problems simultaneously.
«If your only moat is that your product is AI, you're going to be replaced by anyone with lovable in their living rooms.»
— 00:00
«Building is not the bottleneck, and now deciding what to build becomes the bottleneck.»
— 00:08
«AI is not a product in itself.»
— 03:42
«Rather than taking that productivity gain back for ourselves to be able to have more time and brain space... we instead just naturally want to like, oh, I can build twice as much, I should push myself to build three times as much.»
— 05:33
«Don't try to be anyone else. Just try to be yourself and that's what will make your talks or your presentations good.»
— 15:53
«I actually controversially probably think that remote working is really having a big impact on this.»
— 16:29
«We spend way too much time at work for it to be miserable.»
— 17:29
«You have permission to be silly.»
— 19:18
«the more that we try to do things right and use the product language, actually the more alienating we are.»
— 22:00
«we have a new shiny thing. Want to take a look?»
— 23:53
«basically everyone thinks that they can just build what you have built with cloud code in a weekend.»
— 25:08
«my mom doesn't want to try to figure out cloud code and how to how to connect to things.»
— 27:33
«positioning yourself as AI first is actually a downside.»
— 28:10
«I can't predict the next 5 months, much less the next 5 years.»
— 29:09
«I found it simultaneously incredibly inspiring and incredibly depressing.»
— 32:24
«suck out what was special, strip it for parts and leave shells of companies that we once knew and loved.»
— 32:57
Reception
No comments are available to gauge audience reception.
A wide-ranging conference-recap interview that maps Emily Tate's read on the current AI moment in product management — skeptical of 'AI as moat,' focused on positioning and customer value, and threaded with reflections on workplace joy, public speaking, and corporate governance.
34:58