I know I’m gonna get a lot of shit for saying this out loud, but after digging into the research, it is what it is. https://x.com/i/article/2110746268347891712
Being backed by YC doesn't make you special.
Look at you. Fresh YC backing and a 14-part launch thread about how you're changing the world. Congratulations. Now, what does your startup actually do?
News flash. It's 2026, and being YC-backed doesn't mean shit on its own anymore. I could throw a stone in San Francisco and probably hit three YC founders.
Don't get me wrong, they have produced some of the best companies in the world. Airbnb, Stripe, Coinbase, Reddit. I have enormous respect for what they've built. But lately, between the copycat startups, the AI wrappers, and the painfully self-important founder content, I've started wondering what exactly we're supposed to be celebrating.
So yes, I have a bone to pick with YC.
Firstly, everything is a copy of a copy of a copy (in the worst way)
Remember PearAI?
In September 2024, a YC-backed founder announced that he'd quit his $270,000 Coinbase job to build an AI code editor.
Very inspiring, until developers started looking at the code.
Turns out, PearAI was largely built on Continue, an existing open-source AI coding assistant.
The fun part? Continue was also YC-backed.
The less fun part was that PearAI had replaced the original open-source license with its own license, which the founder admitted he'd written using ChatGPT.
Yes. An AI startup used ChatGPT to write the license for a product built on another startup's code.
The founders eventually apologised and fixed the licensing issue. Even Garry Tan later admitted he'd defended PearAI without knowing all the facts.
But PearAI wasn't an isolated example of overlapping products.
A TechCrunch investigation in November 2024 examined nearly 5,000 YC companies and found more than a dozen AI coding startups backed between 2022 and 2024.
Continue, PearAI, Void, EasyCode, Ellipsis, Cosine, Greptile. Some were in the same batch, working with the same YC partner.
And the overlap wasn't limited to coding. The research found similar clusters across payroll, restaurant software, and other markets.
YC's defence is that it invests in founders, not ideas. Garry Tan has argued that more choice and more people building products are good things.
Sure, I get that, nobody owns an entire market just because they got there first.
But surely there's a point where you have to ask what the next startup is bringing that the previous six didn't.
I can understand two exceptional teams independently attacking the same problem. What I find harder to understand is an accelerator celebrated for backing unconventional founders repeatedly funding companies whose pitches sound interchangeable.
At some point, I'd like to see something other than another variation of a product we already have.
Secondly, does the world need MORE GPT wrappers?
I swear every other startup announcement now sounds like somebody took an existing workflow, added a chat interface, and put “AI-native” on the landing page.
AI for meetings, AI for sales calls, AI for documents. AI for managing other AI agents.
And the pitches are getting more ambitious than the products.
Take Nomi, from YC's Spring 2025 batch. Its real-time sales copilot listens to calls and tells salespeople what to say, when to ask questions, and when to close.
Or Den, also from Spring 2025, which initially pitched itself as “Cursor for knowledge workers,” bringing AI agents into a Slack/Notion-like workspace.
I'm not saying these companies are useless. Maybe they're solving real problems. But reading some of these pitches, I keep wondering what happens when the model providers themselves ship the same capabilities.
In fact, Den got so much criticism for supposedly being a wrapper that co-founder Justin Lee wrote an entire LinkedIn post defending the product.
He listed the agent builder, databases, scheduled agents, integrations, templates, and other things they'd shipped.
Fair enough, those are real product capabilities, and they deserve to be evaluated.
But it says something about the current AI startup market that founders are having to explain why their products are more than ChatGPT with extra buttons.
Look, I don't think building on top of OpenAI, Anthropic, or Google is inherently bad. Plenty of excellent businesses are built on somebody else's infrastructure.
I just want to know what you're adding that matters.
What happens when the underlying model gets better? Why will customers keep paying you? How much of your product is genuinely difficult to recreate?
A working demo used to be impressive because building software was hard. Now a small team can make something that looks fantastic in a weekend.
That's great for experimentation. I'm just not sure why we're treating every convincing prototype like the next billion-dollar company.
And please, can we retire the demo where somebody types a sentence and an AI agent performs a task that takes a human thirty seconds?
I've seen enough.
Thirdly, your sloppy LinkedIn content is getting out of hand.
Can we talk about the content for a second?
Because apparently getting into YC now comes with an obligation to publish your entire emotional journey on LinkedIn.
I quit my six-figure job.
Moved to San Francisco.
Got rejected.
Applied again.
Got into YC.
Here's what the experience taught me about resilience.
Brother, I just want to know what your software does.
I'm all for founders sharing their journeys. But sometimes the startup feels like a supporting character in the founder's personal brand.
I came across a LinkedIn post from a founder announcing his YC acceptance. He talked about how exciting it was to build in tech, how legendary YC was, and how grateful he felt.
One of the comments simply asked:
I laughed.
Imagine announcing that you're building a startup and leaving people wondering what the startup actually is.
And it's not just founders. Sometimes YC's own promotional content makes me wonder who's approving these things.
Remember Optifye?
In February 2025, YC posted a demo of the AI-powered factory monitoring startup. The product used cameras to track worker productivity.
The video showed a supervisor confronting an underperforming worker identified as “Number 17,” telling him that his efficiency was just 11.4%.
The internet absolutely hated it.
People criticised the demo for its depiction of worker surveillance, with one commentator describing it as “sweatshops-as-a-service.”
YC eventually deleted its promotional posts from X and LinkedIn.
I watched this whole thing unfold and kept wondering how nobody thought the marketing might backfire.
Again, I'm not against building in public. Quite the opposite.
Tell me about the customer who made you reconsider your product. The engineering decision that took weeks to get right. The feature nobody wanted. The thing you shipped that actually worked.
Those are interesting stories.
But please spare me the fifteenth post about how drinking coffee with ambitious people in San Francisco changed your perspective on life.
It's okay to build a company without turning every Tuesday into a founder lesson.
Fourthly, YC-backed != a good product.
I think we've forgotten something incredibly basic.
YC is an investor. A very successful one, but an investor nonetheless.
Getting accepted means some smart people believe your company has potential. It doesn't mean the product is good, customers want it, or the business is going to work.
Yet somehow, we've turned those two letters into a quality certificate.
A YC founder announces a new product and people immediately want to try it. An equally capable founder without the badge can spend months trying to get anybody to notice.
And I understand why. YC has earned its reputation. Investors, journalists, and customers have limited time, so they naturally pay attention to companies that have already passed through a respected selection process.
But that still doesn't tell me whether the product is worth using.
YC's own motto is “Make something people want.”
I wish we'd pay more attention to the second half.
Because customers are much less interested in who funded you than people on startup Twitter.
If your product is slow, they'll complain. If it doesn't solve their problem, they'll leave. If a competitor does it better, they'll switch.
The orange badge isn't going to fix any of that.
And I feel we're giving far too little attention to founders building excellent businesses without a famous investor behind them.
The ones spending years figuring out difficult problems, getting customers to pay, and building something profitable without ever announcing an accelerator acceptance.
I'd love for us to get just as excited about those people.
All that being said, I do love some of you guys.
Before somebody sends this to every YC founder they know, I should say there are some seriously impressive companies coming out of that ecosystem.
I love what PostHog is building. Its product covers analytics, feature flags, session replay, experimentation, and more, and the company has made an enormous amount of its internal thinking public.
Go through PostHog's handbook. It explains how they work, make money, decide what to build, and structure their teams. Its careers page now says the business generates over $50 million in annual revenue.
I can spend hours reading that stuff. There's more to learn there than in a hundred generic founder threads.
Then there's Supabase.
It took a very real developer problem, making PostgreSQL easier to use, and built a platform that developers now use to ship entire applications.
In June 2026, Supabase announced a $500 million Series F at a $10 billion pre-money valuation. It also released an early version of Multigres, its open-source system for scaling Postgres.
That's a company with a clear product, a huge developer community, and a difficult engineering problem to keep solving.
I'm also a fan of Browser Use. Its open-source project lets AI agents interact with websites, and its YC profile reports more than 110,000 GitHub stars.
You can look at the repository, run the code, and see what the team has built. That's a much more interesting introduction to a startup than its investor list.
And Vapi has a great story. In May 2026, TechCrunch reported that Amazon Ring had evaluated more than 40 voice AI vendors before choosing Vapi to handle its inbound calls, 40 competitors, One customer making a real buying decision. That's far more impressive to me than another funding announcement.
One more company worth mentioning is Bolna AI, which is building voice AI infrastructure for Indian businesses. Handling calls in India is genuinely difficult. People switch languages mid-conversation, accents vary, background noise is common, and businesses are extremely sensitive to costs.
In January 2026, Bolna reported more than 1,050 paying customers and over 200,000 calls a day. Those are company-reported numbers, but they describe a business doing real work for real customers.
And yes, several of these companies build on existing models or technologies, too. That's not the contradiction someone might think it is.
I'm interested in what they've actually built on top, whether it solves a meaningful problem, and whether customers keep using it.
These companies are part of why I still respect YC. Clearly, it can identify and support exceptional founders.
Which is also why I think it's fair to expect more from it.
Anyway, sorry if this article offended you.
I have nothing against founders who get into YC. It's difficult, and you should celebrate
But I think we've become a little too generous with the applause.
Getting funded is exciting. Building something people love is much harder. And somewhere along the way, we started treating those two achievements as though they deserved the same recognition.
Maybe I'm being too harsh. Maybe some of the startups I've criticised will turn into enormous companies, and I'll look like an idiot for writing this.
I'd actually love that.
So if you're a YC founder reading this and think I'm wrong, send me your product. Tell me what you've built, who's using it, and what you've figured out that others haven't.
I'll happily change my mind.
But please don't start by telling me you're YC-backed.
Tell me why anyone should care about what you've built.