Jeremy Jones
speaker
188 appearances
1 recordings
1 series
first heard Feb 2026
last heard 19 Feb
Jeremy Jones’s voice in public audio — every appearance, attributed to the second.
Trend
recordings per month · last 12 monthsRecordings per month over the last 12 months — 1 in all, peaking in Feb 2026 with 1.
Appearances
Because AI is an umbrella term, but just because you're using AI doesn't really mean much to me.
Because what AI actually is, is algorithms.
And algorithms require two things.
They require data and they require instructions.
And when you talk to a general person on the street and you say, hey, have you used AI?
90% of people will think of something like ChatGPT or probably Gemini, right?
Those are the two largest systems that consumers use.
The biggest problem is...
It's like the heuristic I always give to folks is if I ask my three-year-old to make me cookies and I could give her the best ingredients and I could give her the best recipe, she would still not be able to make great cookies.
But where the magic really happens, and this is kind of where we come in, is when you have somebody who is a Michelin star trained chef, you give them the best ingredients in the world, you give them the best recipe in the world, and they know exactly what to do with it.
So that's really how we leverage these systems.
We're very similar to Palantir in a way, where we're not necessarily building our own large language models.
I think that that starts to become a commodity as these systems start to reach natural parity.
I think give it a few years and the capabilities for Claude, Gemini, and ChatGPT will all be more or less the same.
how you use those systems and you apply those systems is really going to be the differentiator.
And really what you need when you're using these systems is really robust data and a really robust recipe or prompting, right?
So when we're prompting these systems, our prompts aren't paragraphs long.
They're usually pages long.
The instructions were specifically what we're looking for, specifically what we wanted to do.
And then the data set isn't just like,
Showing 61–80 of 188 · page 4 of 10
← Previous
Next →