Liftoff with Keith Newman: For Founders looking for a new growth playbook and further inspiration. · Former Journalist, Startup + GTM exec Keith Newman

Agentify: How AI Agents Are Reshaping Companies, Careers, and the Future of Work | Michael Palmer

·31 min·1 clip

Michael Palmer says today’s agents are “agents without agency” that still wait for prompts.

1. Liftoff with Keith Newman centers this episode on AI agents and how they could reshape companies, careers, and the future of work. 2. Keith Newman interviews Michael Palmer, CEO and Chief Scientist of Talos Research Corporation and author of Agentify: The Art, Science, and Engineering of Successful AI Agents. 3. The core question is whether AI agents will stay as prompt-driven tools or become systems with real initiative inside companies. 4. Palmer says he moved from Silicon Valley robotics and startups to a CTO role at a large bank before founding Taos after the late-2022 GPT moment. 5. He says his coding workflow changes every month or every two to three weeks because of new models, tools, and agents. 6. Palmer discusses “vibe coding,” naming Lovable, Replit, Bolt, and Cloud Code as examples of tools that made rapid prototyping feel easier. 7. He argues that fast prototypes still have bugs, security issues, and quality problems, which is why he wrote a chapter called “Beyond Vibes.” 8. Palmer says many AI products will shift from buttons and knobs toward more service-oriented agents that do more of the work for consumers and enterprises. 9. He says specialization and deep customer knowledge matter because big AI companies can “steamroll over” startups and copy features quickly. 10. Palmer says the next major change is adaptability, with agents that take initiative instead of waiting for a human prompt. 11. He says current systems like Claude Code or Lovable can complete a task, but they do not yet understand objectives and continue on their own. 12. Palmer says he does not think current LLMs will reach AGI, even though scaffolding, tooling, and “context engineering” can get systems to solve short, domain-specific problems better and better. 13. He estimates that simulated goals and simulated initiative can get systems “80, 85% of the way there” without true AGI. 14. Palmer says enterprise AI adoption is happening in point solutions first, such as call centers, software development, marketing, and internal tools in Microsoft Word and Google Docs. 15. He says many companies are still frozen, or measuring AI with simple internal usage counts that do not show whether the tools matter to the business. 16. Palmer says software development is farther along than most functions because teams can measure output per developer and cost reduction more directly. 17. The conversation stays direct and practical, with Keith Newman pressing on business models, org design, and how AI changes management layers. 18. Palmer speaks in a builder-to-builder style and frequently moves between product strategy, enterprise buying, and engineering details. 19. Founders building AI products, enterprise operators, and software leaders would likely get the most value from this episode. 20. Listeners looking for entertainment-first storytelling or non-technical chat can probably skip it.

As heard by us

A grounded look at adoption, measurement, and the enterprise barriers around AI agents.

Michael Palmer brings the discussion down to earth by shifting attention from what AI agents might do in theory to what changes when companies try to make them useful in practice.

Read the full review in PlayNext →

Why you'd press play

If you want a practical founder's take on AI agents, this one stays grounded.

Read the full recommendation in PlayNext →

Listen to the show on

Apple Podcasts Spotify

Play clips from this episode More from Liftoff with Keith Newman: For Founders looking for a new growth playbook and further inspiration.