Mikey Shulman

There's something else that goes back to what we were talking about before, which is that if you are interested in a different AI problem, in the problem of how do I align models not to objective truth, but to human taste, I think there's nowhere else to do it.

Yeah, it's really hard. We have the advantage of having a lot of usage. And so we can run, we have a lot of data, we can collect a lot of data, we can run a lot of A-B tests on things. In the future, there's probably a lot more personalization that happens in this domain. But in the meantime, it's similar techniques that align models to human preferences for like RLHF or for chat GPT, etc.

Yeah, it's really hard. We have the advantage of having a lot of usage. And so we can run, we have a lot of data, we can collect a lot of data, we can run a lot of A-B tests on things. In the future, there's probably a lot more personalization that happens in this domain. But in the meantime, it's similar techniques that align models to human preferences for like RLHF or for chat GPT, etc.

Yeah, it's really hard. We have the advantage of having a lot of usage. And so we can run, we have a lot of data, we can collect a lot of data, we can run a lot of A-B tests on things. In the future, there's probably a lot more personalization that happens in this domain. But in the meantime, it's similar techniques that align models to human preferences for like RLHF or for chat GPT, etc.

But it's totally not obvious that that is what the future should be. You know, it's totally not obvious that the same techniques that are used to align LLMs to weird human taste should be the same techniques that we use to align music models.

But it's totally not obvious that that is what the future should be. You know, it's totally not obvious that the same techniques that are used to align LLMs to weird human taste should be the same techniques that we use to align music models.

But it's totally not obvious that that is what the future should be. You know, it's totally not obvious that the same techniques that are used to align LLMs to weird human taste should be the same techniques that we use to align music models.

Totally. We think about this a lot. They will try. I don't know if it will be music. You know, customer support seems honestly bigger. You know, the thing that OpenAI is going after is, let's be honest, I hope to build a giant company. But general intelligence and customer support are both much bigger than music.

Totally. We think about this a lot. They will try. I don't know if it will be music. You know, customer support seems honestly bigger. You know, the thing that OpenAI is going after is, let's be honest, I hope to build a giant company. But general intelligence and customer support are both much bigger than music.

Totally. We think about this a lot. They will try. I don't know if it will be music. You know, customer support seems honestly bigger. You know, the thing that OpenAI is going after is, let's be honest, I hope to build a giant company. But general intelligence and customer support are both much bigger than music.

The future of customer support might be much, much bigger than Zendesk, right? If everything can be automated. You have all these BPOs, you have all these. It's not obvious to me. I'm quite certain that the future of music is also much bigger than it is today, but maybe that's not dawned on the folks at OpenAI. What I will say is that this isn't just, it's the same thing about things being tastes.

The future of customer support might be much, much bigger than Zendesk, right? If everything can be automated. You have all these BPOs, you have all these. It's not obvious to me. I'm quite certain that the future of music is also much bigger than it is today, but maybe that's not dawned on the folks at OpenAI. What I will say is that this isn't just, it's the same thing about things being tastes.

The future of customer support might be much, much bigger than Zendesk, right? If everything can be automated. You have all these BPOs, you have all these. It's not obvious to me. I'm quite certain that the future of music is also much bigger than it is today, but maybe that's not dawned on the folks at OpenAI. What I will say is that this isn't just, it's the same thing about things being tastes.

This is a product game. At the end of the day, people will use Suno because it's a better product, not because there's a better, bigger model under the hood. We are in the business of selling pleasurable musical experiences to people, plain and simple. And so a chat interface is not the be all and end all that people want to do for that.

This is a product game. At the end of the day, people will use Suno because it's a better product, not because there's a better, bigger model under the hood. We are in the business of selling pleasurable musical experiences to people, plain and simple. And so a chat interface is not the be all and end all that people want to do for that.

This is a product game. At the end of the day, people will use Suno because it's a better product, not because there's a better, bigger model under the hood. We are in the business of selling pleasurable musical experiences to people, plain and simple. And so a chat interface is not the be all and end all that people want to do for that.

transformers because it lets us kind of get that to you quicker. We can actually measure. It's super important what you said. And we keep a very close eye on making sure we deliver an amazing experience for your first song because it matters a ton. And I can tell you, for example, we can introduce a little bit of artificial latency into the product and we can see people like it less.

transformers because it lets us kind of get that to you quicker. We can actually measure. It's super important what you said. And we keep a very close eye on making sure we deliver an amazing experience for your first song because it matters a ton. And I can tell you, for example, we can introduce a little bit of artificial latency into the product and we can see people like it less.

transformers because it lets us kind of get that to you quicker. We can actually measure. It's super important what you said. And we keep a very close eye on making sure we deliver an amazing experience for your first song because it matters a ton. And I can tell you, for example, we can introduce a little bit of artificial latency into the product and we can see people like it less.

Yeah, but 10 seconds is worse than eight. I don't see us getting to one second anytime soon, but I think that the crossover that you're talking about, we're still kind of far away from.