“Welcome back to another episode of the Lightcone. I'm Gary. This is Jared, Harge, and Diana. And collectively, we funded companies worth hundreds of billions of dollars right at the beginning.”

This is the opening for the year-end episode for the Lightcone podcast, a regular show with big bosses of the world’s most famous venture accelerator  Y Combinator: President Garry Tan, Managing Partners Harj Taggar and Jared Friedmanm and Diana Hu, General Partner at Y Combinator. I've summarized the 40-minute podcast into this article.

The transformation of AI business models

 The initial skepticism about AI companies has been completely upended. As the partners discussed:

 "The immediate consensus view was all of the value would go to OpenAI. And very specifically, do you all remember when they announced the GPT or the ChatGPT store? The consensus was everything that was built on top of ChatGPT was a GPT wrapper, and the app store was just going to be released and crush every single person trying to build an AI application... It sounds kind of ridiculous to say that now because - Who even remembers the ChatGPT store? Exactly."

They highlighted successful AI applications across different sectors:

"What are the big AI applications today? I'd say outside of ChatGPT itself, the breakout consumer application is Perplexity. The breakout enterprise application is probably Glean, maybe. In legal tech, you have Case Tecks, you have Harvey, prosumer, you have PhotoRoom. The point being, there are many, many applications that have been built not by OpenAI." 

The open-source revolution in AI

 The landscape shifted dramatically with open source developments:

 "The weird series of events where like the weight's being leaked and like Meta just like rolling torrent... That's kind of forced the hand for Meta to launch Llama, which is funny. And people thought, oh, it was just this cool open source model, but it was 18 months behind OpenAI. And people started doing a lot of derivative work out of it. It's like Vicuña and all these other animals related to llamas that came out."

 This led to a significant turning point:

 "The thing that changed from 2023 to 2024 is that during the summer, it was a turning point. It was the first time that the top foundation model in all the rankings, benchmarks was Llama. And that was a shock to the community."

 Multi-model architecture and enterprise applications

A major trend emerged in how companies use AI:

 "Companies started to use multiple models for the applications, like the best one for speed at some point, because sometimes you need to parse a lot of the input very quickly. It's fine if it's a bit more lossy. And then you need the bigger model to handle the more complex task."

 They provided specific examples:

 "A concrete example we gave a couple episodes ago was Camphor. They use the fastest model for parsing PDFs, and the more complex ones, they use O1... Other companies are doing fraud detection, they have this concept of a junior risk analyst where they just use like a fast and easy GPT-4 mini and then they use the bigger one with like O1."

 The evolution of AI development tools

 The partners noted significant changes in development practices:

"2024 was the year that AI coding really broke out. I mean, we had the majority of YC founders now use Cursor or other AI IDEs. They just like exploded over the summer. Devon proved that you could like fully automate like large programming tasks."

This has led to new interview practices:

"Part of the clever interview tricks I've seen is people do pair programming and watch them use the tools and you can really tell if someone really has tinkered with them. It's actually an engineer that is not only good at coding, but also prompting and telling when the AI output is not correct."

Voice AI and vertical applications

 On the potential of voice AI:

 "Voice is just like AI. It touches everything, and there's so many different applications for it that there's probably infinite applications to build where voice is the interesting element of it... things that just spring off the top of your head, like language learning applications... Remote work, like teleconferencing, it's probably like a whole other area."

 They emphasized the importance of vertical specialization:

"Customer support is not really one vertical. There's like many different flavors of customer support and there's like very different on the inside once you get into the details because I think there's very specific types of workflows you need to do per industry... it's just very different workflows if you're building the voice agent to do customer support for an airline, very different than doing it for a bank, very different than doing it for a B2B SaaS company."

AI impact on company building

The partners observed changes in how companies scale: 

"Some people are saying sort of the opposite, which is, I'm going to get my software engineers to write more processes that use LLMs upfront. And, you know, I probably will end up needing to hire that person, but maybe after the series B or C and not right now."

Enterprise adoption and reliability

A significant shift occurred in enterprise attitudes:

"A year ago, one of the things that people said a lot was that these LLMs are not reliable enough to deploy in the enterprise. They hallucinate... That was why a lot of the people said these pilots and POCs won't translate into real contracts... And not only is it translating into real revenue but it's translating into real deployments that are like being used at large scale, you know, doing thousands of tickets a day."

The future of AI development

 The discussion touched on emerging trends:

 "This concept of thinking of AI more as agentic. That is a term that kind of bubbled up a lot this year. It was not in the bubble space of conversation last year. Last year was more about a lot of things that were kind of very chat-like... but now it remixed into a bunch of agents for xyz."

They noted the expanding capabilities:

"The capability of the models keeps pushing in the direction of just being able to do like complex multi-step things. And actually take over your computer and call other applications and perform complex tasks that just didn't seem possible a year ago." 

Growth and scale

The partners highlighted the unprecedented growth potential:

"The time it's taking to reach $100 million in annual revenue is trending down... When they started Andreessen Horowitz, the common understanding was that in any given year, there'd only be 15 companies that year that would even make it to $100 million a year revenue. And they said they ran the numbers the last 20 years. And every decade, the number of companies that could actually make it to $100 million went up by 10x. So what was 15 per year maybe 20 years ago, I mean, we're talking about 1,500 companies a year that have a real shot at actually making that number."

The discussion reveals how AI has fundamentally transformed the startup landscape, creating unprecedented opportunities for growth while continuously evolving in its applications and capabilities. The partners' insights suggest we're still in the early stages of this transformation, with significant potential for further innovation and development across multiple sectors.