Real Alpha
Watch modeSubscribed

Lightcone · 3 min · Jul 11, 2025

AI coding tools cut some experiments from days to an hour

Perplexity made AI coding tools mandatory, and one machine-learning workflow now takes an hour instead of three or four days. Aravind Srinivas also warns that generated code can add bugs engineers struggle to trace.

Aravind Srinivas

Aravind Srinivas is cofounder and CEO of Perplexity.

Held research roles at OpenAI and Google DeepMind before co-founding the AI search engine Perplexity in 2022.

  • Garry Tan, Host · President & CEO, Y Combinator

    Co-founded Posterous and Initialized Capital, an early Coinbase backer.

  • Harj Taggar, Host · Managing partner, Y Combinator

    Co-founded Auctomatic with the Collison brothers and later Triplebyte.

  • Diana Hu, Host · Managing partner, Y Combinator

    CTO of Escher Reality; after Niantic bought it, ran AR for Pokémon GO.

  • Jared Friedman, Host · Managing partner, Y Combinator

    Co-founded Scribd, which YC funded in 2006; has advised 20+ YC unicorns.

About this segment

Perplexity made AI coding tools mandatory, and one machine-learning workflow now takes an hour instead of three or four days. Aravind Srinivas also warns that generated code can add bugs engineers struggle to trace.

From “Aravind Srinivas: Perplexity's Race to Build Agentic Search”, Jul 11, 2025. Watch the full episode

Answers “Is AI paying off for companies?” See all 17 answers

“That is reducing the experimentation time from like three, four days to like literally one hour.”
Aravind Srinivasat 20:58 · Lightcone20:58

Takeaways

  1. Perplexity requires employees to use an AI coding tool.
  2. Cursor can turn pseudocode into implementation, tests and experiments.
  3. Some experiments fell from three or four days to one hour.

In this episode

Full episode
  1. 11:27Google’s incentives leave room for search startups2:48
  2. 17:01A browser can turn an assistant into an agent2:47
  3. 27:33Perplexity sees subscriptions and task fees as revenue streams2:23

  4. 19:48AI coding tools cut some experiments from days to an hour2:33

  5. 23:53Browser agents can work even when sites lack MCP servers3:49Up next
  6. 9:08700,000 New Year's Eve queries convinced Srinivas Perplexity was real2:23

Transcript

Transcribed by machine (Whisper) from the episode’s audio, so a word may be off here and there.

  1. 19:54

    you now have access to code writing AI tools are you Are you guys just like full in on that stuff? Are you vibe coding everything? What's it look like? I mean, you don't want to vibe code everything, right? Like we frequently run into infra issues and you don't want a vibe coder right there fixing it on live things on production. Like I do want like people well-trained in regular software engineering, infrastructure, distributed systems.

  2. 20:21

    Like you don't want to like replace these skills. But yeah, front end design, that's where we are seeing tremendous adoption like Cursor is being used by everybody. We made it mandatory to use at least one AI coding tool and internally at Perplexity it happens to be Cursor and like a mix of Cursor and GitHub Copilot. But yeah, we definitely made it compulsory. And so the way machine learning people are using it, AI people are like, sometimes they read a paper and they can just upload a screenshot of the pseudocode and ask Cursor to like just edit the files to implement this new algorithm.

  3. 20:58

    And then it's able to like write its own unit tests and then run an experiment pretty quickly. That is reducing the experimentation time from like three, four days to like literally one hour. Or like there are people who don't know design and so sometimes I just give them feedback where I take a screenshot of my iOS app and I say this button needs to move here with an arrow and they upload my screenshot to cursor and then ask it to like write a change to the SwiftUI file.

  4. 21:27

    So that level of change is incredible. Like the speed at which you can fix bugs and ship to production is crazy. The more bugs there are, as long as you can fix them fast. Yeah, bugs are always ahead of how fast people can write code, though. But just to be clear, I'm a big fan of all these tools, but it is also introducing new bugs, and many people don't know how to fix them. And they don't even know how the bug got introduced, and they have to go find it again.

  5. 21:54

    So it's not perfect, and I actually like the more newer tools. Claude code seems to be far smarter than what Cursor is able to do. So I'm actually really positive that this is the right future, but there are issues right now. In talking to a lot of the folks here, one of the major questions that I've heard is,