Jaeden Schafer
๐ค SpeakerVoice Profile Active
This person's voice can be automatically recognized across podcast episodes using AI voice matching.
Appearances Over Time
Podcast Appearances
We're not actually, this is a quote from them.
They said, quote, we're not actually on a mission to get as many users as possible.
Now, if they had an insane amount of monthly active users, I'm sure they'd be touting it, but it doesn't seem like it's maybe the strongest
thing about the company.
And so it feels like that's why they might not be sharing it out so publicly.
I think they're trying to kind of reinforce their positioning right now.
They said that they recently introduced a new benchmark for some complex research research tasks.
They're calling it Draco.
They say that their deep research product outperforms things like Gemini.
Perplexity also said that they have built their own AI-optimized search index, which is basically reducing the reliance on third-party APIs.
I think kind of the core thing to Perplexity's strategy right now is that they believe that the future is multimodal rather than betting on a single LLM like ChatGPT or Cloud.
Perplexity is basically routing your question across a bunch of different third-party LLMs, and they pick the best one for answering that specific task.
They actually have a cool feature that I've recently seen, which is like I think it's called like court or something where you basically ask a question to the court and it pitches the question to like all the top AI models.
They all give their response and it has them all deliberate between each other what the best response is.
And then it gives you like based off of Grok, Gemini, ChatGPT, Anthropic.
This is like the best response.
which I think is a pretty interesting, you'd hate to ask a question that one particular model doesn't have a good data set for, right?
Like we know that Anthropic is the best for code and OpenAI might be the best for getting like real world info and Google's the best for, you know, things that need API integrations to a lot of data.
Like they're, they all have these kind of different strengths and, um,
Sometimes I feel like I asked the wrong question to chat.