INSIDE How AI Startups hire, AI Roundtable with Wade Foster, Mikey Schulman, and Ali Ansari | E2225
Timestamps are as accurate as they can be but may be slightly off. We encourage you to listen to the full context.
Podcast Summary
In the first-ever "This Week in AI" roundtable episode, Jason interviews three tech CEOs about the future of AI hiring and talent acquisition. Wade Foster (Zapier), Mikey Schulman (Suno), and Ali Ansari (Micro1) discuss how AI is transforming the tech talent wars, the challenges of competing with massive compensation packages from Meta and OpenAI, and the shift toward hiring globally. (02:30) The conversation explores how AI tools are democratizing development capabilities, making design a bottleneck instead of engineering, and creating entirely new career paths like "human data experts." (24:10)
- Core themes include AI's impact on hiring practices, the emergence of new job categories, training data markets, music industry disruption, and addressing concerns about job displacement and social isolation
Speakers
Wade Foster
Co-founder and CEO of Zapier, a workflow automation platform with 8,000 integrations that has been operating for 14 years. Foster has navigated multiple paradigm shifts in tech and built Zapier into a distributed company with 750 employees, positioning it as a key player in the AI agents and workflow automation space.
Mikey Schulman
CEO of Suno, an AI music generation platform that enables civilians to create production-quality music similar to how Canva democratized design. Schulman leads a team of 120 people, with most employees being musicians themselves, and has positioned Suno as a creative tool that's reshaping the music industry's relationship with AI technology.
Ali Ansari
CEO of Micro1, which started as an AI-powered engineer screening platform but pivoted to become a leading provider of training data for large language models. Ansari's company focuses on high-value verticals like finance, medical, legal, and coding, and he famously pitched Jason's team from outside a Jeni's ice cream shop three years ago.
Key Takeaways
Just Start Doing the Job You Want
The most effective way to get hired in the AI era isn't through traditional applications or interviews, but by actually demonstrating your capabilities. (24:10) Wade Foster emphasizes that high-agency creativity never goes out of style, and candidates should build prototypes, identify problems, and email solutions directly to founders. Mikey shared that Suno hired their first Android engineer who simply built an app and sent it to them, while Ali noted they hired their last two AI engineers who built prototypes of their systems. This approach works because it's easier to build now with AI tools, and it cuts through the noise of traditional hiring processes.
Design Has Become the New Bottleneck
As AI tools make coding more accessible, an unexpected shift has occurred where design is now often the limiting factor in product development. (19:55) Ali explains that Micro1 is experimenting with having engineers create prototypes first, then backtracking to create design files afterward. This represents a fundamental shift in the product development cycle, where functional prototypes can be built rapidly, but the design and user experience refinement becomes the more time-consuming process. This trend suggests that design skills may become increasingly valuable as engineering becomes more democratized.
AI Creates New High-Value Career Categories
The emergence of "human data experts" represents a new career path where subject matter experts earn significant income training AI models in specialized domains. (33:45) Ali reveals that over 200,000 people in the US are now working in AI training roles, often earning more than their day jobs while working 15-20 hours per week. These experts are getting paid to do things they're already skilled at while helping train models that will eventually assist them in their primary careers. The highest-demand areas include finance, medical, legal, and coding expertise.
Adaptability Trumps Domain Expertise
In the AI-augmented workplace, the ability to adapt and learn quickly has become more valuable than deep specialized knowledge alone. (23:02) Wade notes that the most valuable skill is not how much you know, but how adaptable you are. Senior professionals who embrace AI tools while maintaining their domain expertise are positioned to "do some insane stuff" by collapsing traditional skill stacks. The key is being able to go from zero to 80% competency in new domains rapidly, enabled by AI tools that accelerate learning curves.
Agentic Workflows Are the Current Reality
While fully autonomous agents remain elusive, the practical solution is combining deterministic workflows with strategic agent placement. (1:13:13) Wade explains that pure agents struggle with reliability when chaining multiple decisions, but "agentic workflows" work well in production. These hybrid systems use structured, deterministic workflows for most tasks while deploying agents only where LLM decision-making is truly needed. This approach delivers the reliability enterprises need while still leveraging AI capabilities for complex reasoning tasks.
Statistics & Facts
- Over 200,000 people in the US are now working in AI training roles, spending 15-20 hours per week training models and often earning more than their day jobs. (33:45)
- ChatGPT has 800 million weekly active users, demonstrating massive adoption of AI tools across society. (60:03)
- 100% of developers at Suno now use AI coding tools, up from just 15% about 12-18 months ago, with many non-developers also using these tools. (18:03)
