In partnership with
Hey there,
Welcome to where I write about my journey from a stable Big Tech Software Engineering job to the wild and volatile world of Venture Capital.
Have you ever watched one of those nature documentaries where the big fish just casually swallows the little ones? Well, that's basically what's happening with Big Tech and AI startups right now, and it's giving me serious anxiety as someone who's supposed to be finding the next big thing.
I call it the "vertical integration squeeze" - and holy sh*t, it's happening everywhere I look.
🐘 When Goliaths Crush (or "Partner with") Davids
Let me walk you through some examples that have been keeping me up at night:
Remember Character.AI, Inflection, and Adept? These were promising AI startups doing cool stuff with their own models. Then suddenly - BOOM - Google, Microsoft, and Amazon swooped in with their "partnerships." And by partnerships, I mean they basically gutted these companies by hiring most of their talent and licensing their tech.
The startups got some cash and fancy press releases about "strategic partnerships," but what really happened? They were effectively neutered - redirected to focus on narrow applications while the big guys took their tech and talent to build competing products in-house. It's like watching someone get mugged, but the mugger leaves them a $20 bill and calls it a business transaction.
Then there's the OpenAI/Microsoft and Anthropic/Amazon/Google deals. $10 billion here, $4 billion there... sounds great until you realize what's happening: the cloud giants are locking up the top AI research labs with exclusive deals. If you're a startup trying to train your own cutting-edge model, good f*cking luck getting the same cloud pricing that these favored partners get. The playing field isn't just uneven - it's a goddamn cliff.
And don't get me started on the "wrapper" startups that built on ChatGPT's API. Remember all those tools that added features ChatGPT didn't have - like PDF reading, web browsing, etc.? Well, OpenAI just casually added those features to ChatGPT itself, instantly undercutting dozens of startups. Jasper AI went from a $1.5B valuation to layoffs and a down round. Ouch.
Is this just how the tech world works now? Are we all just building features that will eventually be absorbed by the giants? Am I wasting my time looking for the next unicorn when the deck is so stacked against them?
AI founders: Got a game-changing startup?
Building the future of AI? Here's your shot at $50K to make it happen.
The Next Big AIdea pitch competition is live. If you've got an AI startup that's changing how businesses grow, we want to see it.
Record a 60-second pitch and you could win $50,000 cash, $25K in AWS credits, 600K Clay credits + Pro Plan, and exposure to millions through HubSpot Media's network.
Five finalists get flown to San Francisco to pitch live at INBOUND 2025 in front of 1,000+ industry leaders.
Your AI idea deserves more than just another LinkedIn post.
🛡️ How to Not Get Crushed (Maybe)
After a lot of late nights thinking about this problem (and several existential crises), I've come up with some strategies that might - MIGHT - help startups navigate this minefield:
Find an actual moat (not just wishful thinking). If you're building on someone else's tech, you need something proprietary ASAP. Unique data sets, network effects, or specialized industry knowledge that the giants don't have and can't easily replicate. One founder I advise is building an AI legal assistant with a massive corpus of annotated legal briefs - data that generalist AI companies don't have. That's a real moat. A pretty UI on top of GPT-4? Not a moat.
Diversify your dependencies. Being completely dependent on one cloud provider or API is like putting all your eggs in a basket held by someone who might decide to drop it (or charge you 10x more to hold it). Always have a Plan B - whether it's alternative providers or open-source models. And for god's sake, negotiate hard when partnering with big firms. I've seen startups sign away their future for short-term gains, and it never ends well.
Avoid the elephant's favorite spots. Be realistic about what Big Tech considers their core territory. If your pitch is "We're building a better general AI chatbot than Google," I'm probably not investing. But if you're focusing on a specific vertical that's too niche for the giants to care about (at least for now), you might have breathing room. The film post-production workflow? Healthcare compliance? Those corners could be your sanctuary... for a while.
Move faster than they can. Your one advantage as a startup is agility. By the time a giant internally approves copying your feature, you should be two innovations ahead. Stay obsessively focused on customers and iterate quickly. When the big competitor emerges, pivot to where they're weaker. It's exhausting, but it's doable with the right team.
Support open ecosystems. The more we all support open-source models, open standards, and interoperability, the less power the giants have. As a VC, I'm increasingly interested in backing companies building alternative infrastructure - from open-source AI models to independent cloud optimizations. The industry doesn't want a future where two companies own all AI progress (well, except those two companies).
Be careful with "strategic" money. If a Big Tech venture arm wants to invest in your startup, proceed with extreme caution. Yes, it's validating, but are you giving them a peek into your roadmap? The ideal scenario is when your goals truly align with the big partner - otherwise, that "friendly" investor might later become an acquirer who dictates terms.
🚨 We Need to Keep the AI Playing Field Open
I'm not usually one for soapbox moments, but this matters. The future of innovation depends on new ideas from new players, not just whatever suits a few giant corporations.
For founders: Stay informed and be strategic. Use the giants' platforms, but don't become entirely dependent on them. Be ready to zig when they zag.
For my fellow VCs: We need to double-click on due diligence around platform risk. And crucially, we should support the "picks and shovels" startups too - companies building open models, new chips, alternative cloud services. These might not have the immediate hype of an "AI app," but they ensure the whole industry isn't beholden to a few gatekeepers.
For Big Tech employees: Use your influence internally to advocate for fair ecosystem practices. Not every feature has to copy/kill a startup - sometimes partnering or letting an API ecosystem flourish is healthier long-term.
For policymakers: Keep pushing for rules that ensure open access. Whether it's antitrust action or funding open research and compute resources, we need to avoid a monopolized AI future.
Innovation in AI should be an open frontier - not a company town where the landlord sets all the rules. Vertical integration is a powerful trend, but it doesn't have to mean doom for startups. With the right strategy, entrepreneurs can still chart successful paths.
The storm clouds are definitely on the horizon, but I'm still cautiously optimistic that the right startups can navigate this landscape. It just means we all need to be a lot smarter about how we build, invest, and grow in the shadow of the giants.
What do you think? Have you experienced the vertical integration squeeze firsthand? Any strategies I missed? Drop me a comment - I'm genuinely curious how others are handling this challenge.
Until next time!
Signing off and signing zero checks,
SWEdonym