Dinner with Casey Winters: 🤯 The Most Important AI Insight No One Wants to Hear

We’re still very early. There’s no “Google/Facebook Ads moment.” And defensibility is shakier than most founders admit — Casey Winters breaks it down.

Last week I had a privilege to host Casey Winters, founder and CEO of SuperMe as part of Icons series.

Some of the most interesting insights from our dinner - about marketing, defensibility, and career strategy in the AI era:

1. Marketing for physical vs. digital products is fundamentally different.
In digital, you can build your way into better marketing. Product, engineering, and even sales can literally fix what used to be marketing problems. That’s why product management has absorbed so much of marketing’s historical power. Marketing was left with positioning and PMM.
In physical products, you don’t get that luxury. If the product is wrong, marketing pays for it.

2. Every major platform shift takes ~4 years before new distribution channels emerge. Casey laid out a fascinating pattern:
- Internet mainstream in 1994 → Google takes off in 1998
- iPhone launches in 2007 → FB mobile ads explode in 2011
- ChatGPT arrives in 2022 → we haven’t seen the new AI-native distribution engine yet.

So the “Google/Facebook Ads moment” for AI probably won’t appear until 2026. We’re still early.

3. GTM must match the value prop.
If your value prop requires customization → lead with sales.
If it’s self-serve → lead with marketing/PLG.
Misalignment here delays everything.

4. Defensibility still matters.
UberEats, DoorDash, Grubhub scaled with negative margins because VCs were betting on network effects and liquidity.
Today, AI companies are scaling incredibly fast — but many with unclear moats.

5. His career advice was refreshingly direct:
Don’t optimize for titles. Optimize for joining the best companies.

At great companies, you get:
• Equity upside
• The chance to learn from the best talent
10 years ago, the winners were obvious — Uber, Airbnb, Snap.
Today? Much harder.

His recommendation:
- Join one of the top 50 AI companies, watch the category from the inside, and if the company isn’t moving fast enough → switch early.
VCs have portfolios.
- Founders can sell secondaries.
- Employees need to be strategic - they have the most to lose.

If history holds, the real defensible winners in AI will become obvious around 2026.

Fascinating conversation and huge thanks to Casey for the perspective.