This video centers on questions that real founders are curious about, with YC (Y Combinator) partners candidly discussing AI business, the timing of pivots, hiring timing, open-source strategy, and various other startup realities. The core conclusion is: "learn fast, find real customer value, and have the founder personally experience everything before making decisions." It is packed with especially realistic advice on entering new markets with AI, tackling technically challenging ideas, and smart hiring timing.

1. Taking on Legacy Industries with AI: What to Start With and How?

As a founder, you face two "magical" problems: who to target, and how to capture customer attention. The first question in the video was: "When entering a traditional industry (e.g., accounting) with AI, with full automation as the ultimate goal, what should you put on the market right now?"

The YC partners introduce three main approaches:

• Sell AI software to accountants
• Start a "full-stack" accounting firm yourself
• Acquire an existing accounting firm and introduce AI within it

"The most common approach is the first one. The majority of actual YC startups start this way."

In the early stages, rather than automating the entire accounting field, they advise that it's effective to automate only a core area that can be built quickly and provides value and pitch that with AI.

The second approach (starting your own firm) risks the organization growing too fast as manual work remains less automated.

"As your automation rate increases (e.g., 20%, 30%, 50% of total), track that metric consistently. If you expand headcount too early... you'll end up hiring 20-30 accountants doing manual work. The real metric is the growth of automation rate. Not revenue."

The third approach (acquisition) is good for acquiring entry customers but is uncommon due to the difficulty of changing existing culture.

In the early stages of an AI startup, the emphasis is: "Find early partners (experts, accounting firms, etc.) who are actively helpful right now and genuinely interested in adopting new software."

"These customers are roughly at the 'earliest early adopter' level in the Crossing the Chasm model -- they're not easy to find."

2. How to Choose Your Market: Mid-Market vs. Enterprise, Balancing Growth and Defensibility

Next came the question of whether, in areas where customer acquisition and growth are slow like AI enterprise products, "if investors want fast growth, should you target SMBs or the mid-market from the start?"

The YC partners say that speed of learning is the most important thing for a startup, and rather than targeting large enterprises right away:

• They recommend starting with smaller segments, customer groups where you can move more nimbly.
• This way, the feedback loop turns faster and you can quickly learn about product/sales/user issues.
• However, if the problem only concerns large enterprises, they suggest trying at least the smallest company that 'has the problem.'

"The biggest challenge for startups is whether customers truly feel the pain and how quickly you can learn and respond."

They add that pre-qualifying customers as decision-makers with real motivation is just as important as segment selection.

3. AI Employees vs. Real Employees? Hiring and Growth Solutions for Early Startups

The thorny question of "AI-based SDRs (sales automation), growth hackers, communicators -- who should you actually hire and when?" also came up.

• YC warns early founders not to expect too much from 'AI sales team' solutions.
• They answer that it's faster to sell the product yourself, acquire customers, learn the sales method through massive trial and error, and only then scale up AI/sales personnel.

"Two big magic tricks: 'Who to sell to and how to get their attention' -- figure both out as a founder first, then hire sales and marketing people."

They repeatedly emphasize that in the early stages, founders should personally experience each function (sales, marketing, product, etc.), thoroughly understand the work, and only then begin hiring.

4. Betting on Barriers to Entry and Model Growth: Invest Now or Wait for Technological Innovation?

They also address the dilemma: "In the current AI frenzy, should we invest now for even temporary advantage? Or should we wait for the next revolutionary model (GPT-5, etc.)?"

The YC partner says that even if more powerful models come in the future, the experience left by current investment and trial-and-error is never wasted.

"What matters is investing now, experimenting, and learning. When a new model comes out, you can build on top of it from day one."

5. When and Why to Pivot: The Difference Between a 'Good' Startup Idea and a 'Great' One

In situations that are "going somewhat well but not explosively," "when to pivot" is something every founder worries about.

• The YC partners identify real user value and growth speed as the most important signals, firmly stating that just because revenue is coming in doesn't make it a killer idea.

"About our product -- are users truly passionate about it, seeking it out every day? Don't fool yourself into thinking 'some people buy it and it sells well.' Keep testing whether you're actually solving a problem."

Pivoting is an extremely difficult choice, and the founder's energy and inner conviction are absolutely necessary.

Sharing real examples, they revealed that through genuine market conversations and experiments, they discovered a truly valuable new idea and pivoted boldly when 'conviction' emerged.

"After changing the idea, revenue dropped, but there was real conviction in the founders' voices. That can be the signal for a pivot."

On distinguishing between a good startup idea and a great one:

"Truly great ideas are created by people who experiment with everything quickly and relentlessly, doubt themselves, pursue endless improvement, and don't spend too long wondering 'is this great?'"

6. Should You Avoid Technically Difficult Ideas?

To the question of "should you give up if it's too technically difficult or can't be implemented quickly?", the advice is actually the opposite.

• High technical barriers to entry mean no one else can easily take on the challenge!
• YC actually believes the most difficult challenges can make the best startup ideas.

"If it's a really hard problem that everyone else has given up on but you have the courage and capability to dive in, it's the greatest opportunity!"

However, they emphasize the one thing to guard against: the "hidden delusion" of not talking to customers while waiting for the product to be finished.

"We spent 6 months focused only on technology development, but looking back, even with an incomplete product, we should have been alongside customers experiencing the problem firsthand -- we would have grown faster."

If necessary, in the early stages narrow the scope and communicate with the market through a version (MVP) that can be built in a short time (2-3 weeks).

7. When Should You Start Hiring? The Real Timing and Criteria

The question "When and what kind of person should you hire beyond the founders?" also comes up frequently.

The YC partners assert that "the real time to hire is when you're too busy to even do interviews."

"If you have time to think about hiring, it's too early. When there's so much work that a normal work schedule can't handle it all -- that's the hiring signal."

• In the early stages, nearly 100% of hires come through referrals and networks as "opportunistic hires."
• Pushing hiring based on simple metrics (e.g., headcount growth) is dangerous, and "hiring itself is never success," they firmly state.

"The era when hiring equaled success is over. Now it's actually trending to boast great revenue with fewer people."

They also share the practical experience that smart hiring is problematic if done too early or too late.

8. Enterprise SaaS Open-Source Strategy: When, Why, and How?

Finally, they answer: "When is it advisable to open-source an enterprise SaaS product?"

• For developer-facing products, open source has strengths in trust, adoption, and market entry.
• But even in the enterprise space, more companies are strategically choosing open source to increase "customer trust and purchase decision speed."

"In certain industries (healthcare, CRM, etc.), open source becomes an important weapon for 'transparency/trust/compliance.' Even if customers can't actually look at the code, the belief that 'I can check anytime' accelerates their decision."

Additionally, especially in the AI field, demand for self-hosting and open source is increasing due to data privacy and security concerns:

"Previously, when customers asked 'Can we install it ourselves?' it was considered unrealistic, but now many startups are making it happen quickly."

Of course, they don't omit the downsides: management and support burden, and higher pricing.

Conclusion

The gist of this video can be summarized in one phrase: "Growing startups obsess over hands-on learning, rapid education, and confirming real value, and all experiments must be experienced firsthand by the founder." Rigorously validating real customer problems, real market signals, and product growth potential -- then boldly taking on opportunities, and quickly pivoting or scaling up when needed -- that is the energetic message on how to evolve into a truly great startup.

"Finding a truly great idea, experimenting relentlessly, and never giving up until you're convinced. That's the hallmark of the best founders YC meets!"