Eric Woo

In the span of three weeks (and one long trip to Asia/China), I’ve gone from eager AI builder to AI user (playing catch up). The reality is that the ROI as a builder is incredibly low given how fast the use cases with moats evaporate. Instead, maximize your productivity in the AI agent candy shop.

Used to be good pattern recognition led to being early to the trend and scaling while the world catches up. Today, AI is making pattern recognition nearly instantaneous, which means scale is commodity (VCs get the memo?). Not going say that everything now shifts back to networks and human relationships—they have been, and always will be, a huge factor. Rather, what’s left is business model innovation. Taking and rolling up AI-powered unit economics across a long tail. Or being the picks and shovels to the new generation of AI SMB. This definitely is not VC and while it smells like PE, that take is a bit antiquated. Lurking here is some form of a new asset class…

Q. What do you get when you put three AI agents together in a chat? A. Nothing. Until you give them instructions.

AI agents and token usage rn is basically like your kids brushing their teeth with the faucet running.

Imagination and nuance is your best moat against AI. Turning ideas into reality is pretty much a forgone conclusion. Ideas still require understanding of human need and value. Nuance is the refinement layer that determines who’s first and who’s second/last. Founders and VCs (those at the forefront of idea generation) are supremely positioned—assuming they are ingesting the latest AI developments. Speed wins. AI is killing the “how does it work?” mentality. “Fail fast” is no longer a cheeky euphemism, it’s become the operating handbook. I am both so sacred and so excited at the same time…

The mass proliferation of AI awareness rn feels just like the early junk mail/spam phase of the email era. Going to sit this out a couple months until content starts feeling fresh and curated.

AI can do the work, but can it do the sale?

In 2008, I had a short stint leading search engine marketing (SEM) for a fintech startup. With the rise of agentic knowledge creation and the attention economy, the rebirth of SEM is coming. This is different from current SEO rebirth because people WILL NEED to $ for agentic or human “eyeballs.” Huge opportunity.

When I started vibe coding (just three mo. ago), it felt like sitting next to a bubbling brook with the tranquil sound of free-flowing water. Today, vibe coding feels like:

Giving “the people” AI is like giving the “caveman” fire. All sorts of crazy shit is going to happen.

This is the year we go from “AI for X” to “X for AI”. Anyone building AI wrappers is on notice. Management, orchestration, etc. of AI agents is where it’s at. Heck, wouldn’t be surprised to see a tool that will mass ➡️ agents to 🗑️.

Pro tip for LPs: VC is a participation (not spectator) sport. It will take 8-10 years to get bulk of money back. If you are on sidelines, that’s a long wait. But if you learn, gain strategic value, and compound your access thru the relationship economy, it can be very rewarding.

Entrepreneurs have a burning, incessant desire to be proven right or proven wrong. Everyone else simply gives up way earlier.

For Emerging VCs, next week (Raise and Equal Summit) will mark the start of a new cycle. If you thought you were already running a marathon, that was the 10k part. Time to decide: tap out, press on, or pause to train for marathon. Choose wisely, but choice must be made.

Distilling my experience as a VC allocator: “Early stage requires conviction, late stage requires consensus.” You want the dissenting opinion pounding the table at early stage. At late stage seek diverse perspectives so as to not be blinded by own views. Hire appropriately.

Tip for fellow startup CEOs: customers PAY YOU to solve their problems, YOU DON’T CHARGE customers to justify the work you do. Let that one sink in…#whatsyourvalueprop

It’s time to cast away previous notions about the fundraising playbook in VC. Being a “translation” layer between GPs and LPs requires so much more than getting them in the same “room”. Only through data and investment analysis can we connect the capital.

On key lesson from last few weeks as a startup founder/CEO: leadership worries about the problems so employees don’t have to. Then by charting the strategic direction, employees can then worry about the solutions. Such a fine line between chaos and focus.