"The Turing test has passed, and no one seemed to care."

On October 17, 2025, Warp CEO Zach Lloyd joined the No Priors Podcast to explore how agentic development tools will reshape software creation, why intelligence and consciousness remain separate concepts, and what automated engineering means for teams.

Distilling Intelligence Without Consciousness

Lloyd framed recent AI breakthroughs as an exercise in extracting narrow capabilities rather than recreating people. “The way I think about the advances is we are distilling intelligence,” he said. “It’s fascinating how much intelligence you can get out of just next-token prediction.”

He cautioned against conflating performance with sentience. Today’s systems can execute tasks we recognize as intelligent, yet remain mechanistic. “We have distilled something that can do things we recognize as intelligence, but it’s totally mechanistic,” he explained. Referencing the sci-fi novel Blindsight, Lloyd argued that intelligence and consciousness can diverge.

That perspective led him to a surprising conclusion: “The Turing test has passed, and no one seemed to care.” Because we understand the underlying machinery, he believes society hesitates to attribute consciousness to AI even if it mirrors human conversation. “As long as we know what it’s doing, we’re very unlikely to attribute true consciousness to it,” he said.

From Prompts to Automated Development

Turning to the developer experience, Lloyd charted a three-stage evolution. For most of his career, engineers “developed by hand.” The industry is now transitioning to “develop by prompt,” with “automated development” on the horizon.

He rejected the narrative that engineering expertise is losing value. “I really don’t think engineering expertise is going to become devalued,” he said. “It’s more important to know what you’re doing as an engineer than it ever has been.”

Warp views AI agents as junior teammates that still need experienced oversight. “If you didn’t have someone senior watching them, you end up in a situation where these agents will make code that creates bugs or security issues,” Lloyd warned. The result is a premium on senior engineers who can architect systems, review code, and keep complexity in check.

Owning the Front Door for Developers

Lloyd also weighed in on competitive dynamics between developer tools and large model providers. While companies like OpenAI, Anthropic, and Google can bundle coding assistants inside their platforms, he argued that developers still start their workflows in local applications. “The front door is still a native app that someone downloads on their computer,” he said. “That would be the terminal or the IDE. Controlling that is a really interesting front door.”

He hopes increased competition at the model layer will commoditize access to powerful AI, enabling Warp to focus on workflow design. “If at some point the models are good enough where coding is solved, maybe they have an advantage from brand and scale, but I think the advantage is not as entrenched,” he noted.

Looking ahead, Lloyd is most excited about programmable agents that eliminate repetitive work altogether. “Allowing developers to automate parts of their job that they don’t like doing is a big capability,” he concluded. “Automation is a better place to be than productivity enhancement because it’s easier to prove the ROI.”