Entrepreneurs Are Rushing to Use AI. Here Are 8 Questions You Should Ask First.

As generative AI sweeps the culture, entrepreneurs have been racing to seize market opportunities. But as ChatGPT quickly gets smarter, new questions and possibilities emerge. How do you stay on top of it all — or even begin? Here's how leading investors, founders, thought leaders , and tech luminaries break it down, in eight simple questions.

---

---

1 . Where do I start?

No matter what industry you're in, there's probably a generative AI tool specifically crafted for your operations — so these are the first resources you should seek out, advises Anna Barber, partner at M13, a VC firm that invests in early-stage tech startups . If you have a large team of developers, you can experiment with the text-to-code assistant GitHub Copilot. Designers should be testing tools like Midjourney, which creates images from text prompts, and creators should be trying storytelling aids like Tome or HyperWrite. For searching large sets of complex data, Humata, an M13 portfolio company, is powerful and easy to set up, Barber says. These tools help startups speed to market and do more with less.

Next, think about how AI might enhance your product, whether you're a telehealth platform adding chatbot features that match patients to therapists or a legal firm that could use AI to search volumes of cases. "In the future, every software product will incorporate AI," says Barber. But AI features need to be evaluated and prioritized like anything else on your product roadmap — not simply added as an afterthought. Ideally, you should be building AI into what you are already doing.

2 . What are the best ways to recruit AI talent?

Competition is fierce. LinkedIn has seen a significant rise in demand for AI-related jobs in the past year, with a 21-times increase in global English-language job postings that mention GPT or ChatGPT. And in the past five years, it's seen a near tripling in the number of LinkedIn members who have "Head of AI" job titles.

To find the right talent for your company, networking is key. Surround yourself with people you want to hire, and then beta test what it's like to work with them by teaming up at a hackathon, says Chappy Asel, cofounder of GenAI Collective, a community of thousands of developers in the San Francisco Bay Area. Many universities sponsor these types of events, or you might find a local collective of AI and data science professionals who meet regularly.

If you're in Silicon Valley, you can also scope out AI talent at hacker houses, where cohorts of founders live and grow their startups together in dormlike settings. AGI House , for example, is set in a sprawling multimillion-dollar mansion where it hosts networking events featuring speakers ranging from Google cofounder Sergey Brin and OpenAI's Andrej Karpathy to pop star Grimes . "We have multiple startups launching from our hackathons," says AGI House founder Rocky Yu. And if you're specifically looking for a business partner, Y Combinator (the accelerator that launched OpenAI) has a free online Startup School that offers founder matchmaking.

3. Can ChatGPT boost my sales?

Most people think of ChatGPT as a tool for external communications: It writes some marketing copy, you refine its work, and then you share that with your customers. But ChatGPT can also help you refine ideas internally, says Conor Grennan, head of generative AI and dean of students at NYU Stern School of Business — especially if you engage it in role play.

"Ask ChatGPT to be a difficult customer who pushes back on every pitch," he says. "It's a great way for your sales teams to learn how to have hard conversations. You can then ask how you could have been more persuasive." Or tell the bot about your product and prompt it to create different pitches to radically different kinds of clients, and ask who else you might try to sell to.

Another helpful exercise: After telling ChatGPT about your product, target audience , and sales strategy, ask it to act like a "consultant" to poke holes in that plan. After it answers, have it switch roles and create solutions for all the problems it just brought up. Next, ask it to poke holes in those solutions. As you keep going, says Grennan, ChatGPT will hone its responses until you're able to get at the heart of the issue and the best solutions for it.

4. Should I use a tool like ChatGPT, or build my own model?

Right now, there's no perfect answer. "Startups training their own models are gaining the highest valuations, even for those focusing on niche use cases like legal and customer support," says Brendan Burke, senior emerging technology analyst at PitchBook. "But this strategy can incur high cloud costs and violate privacy agreements made with customers for their data, and can be risky without significant VC funding ."

On the other hand, using off-the-shelf models can also be pricey and require extensive customization. If you're building a startup with AI at its center, Burke says that the best approach might be to build an MVP of your product that uses a publicly available tool like ChatGPT and cloud resources available at platforms like Snowflake and Databricks. Then tell the story about your greater ambitions. "Investors are willing to fund plans for new model training if unique data is available and the costs are well-understood," he says.

Keep in mind that even the available options have issues. Currently, more than a dozen bestselling authors, including George R.R. Martin and John Grisham, are suing OpenAI in a class-action lawsuit for allegedly training its AI models on these authors' works without permission. "We often debate, do we build our own model? Or do we go with another model and use the data that we have to train on top of it?" says Marissa Mayer, former CEO of Yahoo and now cofounder of the consumer apps startup Sunshine. She's hedging her bets — building smaller models for certain datasets while using larger public models for others. "You protect yourself both ways," she says.

Image Credit: Create Labs

5. Can I beat the tech giants with my AI product?

Yes, according to FPV Ventures cofounder and managing partner Pegah Ebrahimi. "Artificial intelligence," she says, "can level the playing field."

With the proliferation of powerful new AI applications and infrastructure tools, barriers to making products have never been lower. With the ability to incorporate AI directly into tech stacks and workflows , startups can outdo enterprises when it comes to execution. "The question today isn't, 'Are you an AI company?' but, 'How are you using AI to make your product indispensable?'" says Ebrahimi. Some of the most successful companies that took advantage of the mobile wave were those that couldn't have thrived or even existed before cell phones, like Uber. Think of using AI in the same way.

To prevent getting "Sherlocked" by a giant tech player introducing a feature that renders yours irrelevant, GenAI Collective's Asel says one way is to niche down. For example, take construction management and build a competitive moat with market-specific data like zoning laws, localized cost estimates, blueprints, permits, and project schedules — unique datasets that, when fed into your AI model, will significantly enhance its predictions and decisions specific to construction tasks. That way, the model will become smarter and more valuable as it's used.

Whatever you do, Ebrahimi urges founders to make bold bets and imagine revolutionary improvements that weren't possible before. To beat large companies that tend to move more cautiously, it's critical to be agile and iterate quickly. "You have less to lose by trying things out," she says, "so use it to your advantage."

6. What kinds of AI startups are venture capitalists looking to fund?

Despite the AI hype, VC cash is not as free-flowing as it was two years ago — and with a rush of new AI startups coming into the space, it's getting harder to secure funding. The key to beating the odds, say several VCs, is your data.

"The billion-dollar question is, 'Where are the moats?'" says John Cowgill, partner at Costanoa, who leads investments in applied machine learning and deep tech. The answer for him is proprietary access to data that no one else has or can replicate. SignalFire's CEO Chris Farmer agrees, saying they're "looking to fund AI startups building up defensibility through unique datasets that competitors and OpenAI lack."

Both VCs add they want to see AI products that improve with interaction: "Did a user accept the recommendation from the AI system? Did they take a next-step action, or bounce out?" says Cowgill. "Smart AI entrepreneurs are thinking about how to leverage this metadata to make their application smarter with each new user they acquire."

Another way to get attention from top investors is to build in a space that hasn't seen much AI-related innovation yet, like supply chain, manufacturing, and big corporate functions like accounting and legal, suggests Radhika Malik, partner at Dell Technologies Capital. "Then you want to show that you have some unique understanding of the problems that can be solved with AI in that specific space," she says.

7. Once I have a product, how do I land my first customers?

AI may sound exciting, but it's not always an easy sell — especially to large enterprise clients. That's because AI touches so many departments, and will often involve dealing with IT, security, legal, product line, compliance, and data and analytics. Budgetary authority for that is not often clear, says Avivah Litan, a distinguished VP analyst at Gartner Research. Typically, the buyer is the chief information officer or chief information security officer, but you should ask around to find the right person.

As with any sale, make sure your elevator pitch has an easy-to-understand value proposition that solves a problem that your potential client is facing and explains why your solution is better than the competition. Ideally, says Litan, "find a customer with a well-known brand name and work with them to fine-tune your product and service." (This may seem obvious, but many startups don't think about it.) Once you get your first few big clients, the rest will come. "And make your prospects pay for proof of concepts. Don't give away your product," Litan says.

8. Do I need to jump in now, before it's too late?

That depends on the business you're in. Here's what Thomas Tull suggests; he's an early-stage tech investor and chairman of the U.S. Innovative Technology Fund. "A thought experiment you need to run is, How much would I care if my competitor was able to implement [an AI solution], and what impact would it have on my business? " he says. "If your answer is, it would severely hinder or put me out of business , then I would think very seriously about it."

If AI can significantly impact your business, then many investors and founders say you have little time to waste. "I'm still amazed how little founders know about the range of AI tools that exist right now that could unlock value in their company," says Matt Higgins, cofounder of RSE Ventures and an executive fellow at Harvard Business School. "This isn't some distant, theoretical, speculative tech. Startups are the ones that have the most to gain because AI enables a fledgling company to bridge skill and resource gaps." He points to all kinds of things AI can pull off for a fraction of what it would cost even a year ago — business plans, financial modeling, logo creation, content creation, chatbots , and on and on. "Don't wait to be spoon-fed AI innovation," Higgins warns. "By the time it becomes conventional wisdom, your competition will have eaten you alive."

Marissa Mayer agrees you can't afford to sit on this. "AI is probably going to be the greatest inflection point in our lifetime, potentially bigger than the internet. And you can either decide to participate in it," she says, "or there's a question of how relevant you get to be in the long run."