What it takes to build and succeed in the age of AI, from a top tech investor

Thanks to recent developments in AI, it’s easier than ever for a startup to prototype a product, automate workflow, write code, and reach a potential customer. The tools are cheaper, faster, and more widely available.

In a world where more people can build a business, however, how do you stand out?

That question was at the heart of “Building in the Age of AI: What Skills Founders, Operators, & Investors Are Looking For,” hosted by Columbia Business School as part of a16z’s Tech Week 2026. The conversation, moderated by Associate Dean Gracy Sarkissian, brought together Bryan Kim, consumer AI strategist and CBS alum; Professor Mattan Griffel, a two-time Y Combinator-backed entrepreneur; and Jacqueline Deprey ’26, a former CBS MBAxMS student and co-founder.

Together the panel examined what it takes to build a business and lead in the age of AI, concluding that the future belongs to leaders who can “build their way in” by pairing business judgement with hands-on AI-fluency, product instincts, and speed.

Build with momentum

Kim, who invests in AI application companies, says the post-ChatGPT era has changed what he looks for in founders. Now a startup’s competitive advantage lies in its momentum: “There is no moat” separating a promising startup from its competitor — proprietary features, technical complexity, and the time required to build. When rivals can use the same tools to imitate products and ship quickly, the real advantage becomes how fast a company can learn, improve, and reach customers, according to Kim.

Before the current AI wave, he noted, consumer startups often won by obsessing over craft, customer retention, and product nuance. In a mature app ecosystem, founders had to find small openings in the market and build with precision. 

AI has completely altered that calculus, since the market is moving too quickly for founders to rely on a “build it and they will come” strategy, according to Kim. He now places greater weight on a founder’s technical skills and a product’s ability to reach users quickly. 

Product, design, and engineering skills, he argued, collapse into a broader “builder” function. That means a standout founder doesn’t just understand or speak fluently about AI, but also needs to have working knowledge of using AI tools and proof they can turn ideas into working products.

Griffel, who teaches technology and entrepreneurship at CBS, sees the same pattern in the classroom. He noted that even in a world where AI can generate code, business students still need to learn technical coding skills to get ahead. AI, he argued, can narrow the distance between a beginner and an experienced founder, but it can’t replace the judgement required to structure a problem, choose the right tools, or understand what a product should do. 

Turn ideas into evidence

Because AI makes it faster to build products, it becomes easier to confuse activity with real market traction, the panelists pointed out. Speed matters the most when it produces evidence, or real signals from customers.

Kim described how some companies can appear narrow at first glance while addressing a much larger human need. He noted how Function Health, which a16z invested in last year, may look like a blood-testing platform, but the deeper demand it taps into is people’s desire for more control over their health. 

Investors listen for more than a product description — they want to understand the scale of the need, the ambition behind the company, and the founder’s ability to reframe a market.

Founders often spend too much time refining an idea before exposing it to the market, Griffel added. At his company, Ophelia, which expands access to opioid addiction treatment, the team tested demand before it could legally launch care by posting a Craigslist ad asking people struggling with opioid addiction to share their stories. A flood of responses revealed the depth of the problem and gave the founders early insight into what audiences actually needed.

Alignment is another important dimension of building with AI, added Deprey, who is building Perpetua to help families access the online accounts of deceased loved ones. The technology can make companies more efficient, but leaders still need to decide where automation fits their values, customers, and business model.

Prepare for traditional jobs ‘to not exist’

The job market of the future, the panelists argued, will reward employees who can design and manage AI-enabled systems across functions.

It would be wise to prepare for traditional jobs “to not exist,” Deprey said. Instead, use AI to automate lower-value work, increase output, and move toward the next challenge.

Companies want to implement AI but often do not know how, Griffel added, and that creates opportunity in functions from marketing and operations to finance and customer service. Entrepreneurs who can master the art of selecting tools and connecting these systems using AI can find success.

One often underestimated asset that’s on your side, though, especially if you’re a student: time, said Kim. Time gives you the freedom to experiment and learn. 

So, if you have an idea, “just build a thing,” Kim said. In a market where execution is increasingly visible, that may be the clearest way to stand out.