Use AI Like It's Already Superhuman | Felix Rieseberg artwork

Use AI Like It's Already Superhuman | Felix Rieseberg

MTS

October 19, 2026

Felix Rieseberg, engineering manager at Anthropic, discusses the philosophy of human ambition with AI models, the evolution and internal timeline of Claude Code, and how LLM-driven tools empower creative expression and generalist agency. Turn ideas into software people love.

0:01

SPEAKER_1

All right, we are live with Felix Rieseberg, who is a member of technical staff and engineering manager at Anthropic, where he works on Cloud products, including Cloud Cowork, the Cloud Desktop app. Earlier this month, he published How to be More Ambitious, a tutorial for humans with models. So as a human, I would love to unpack this.

So there is a lot of good stuff in this article, but I guess high level. What was your view? You talk about ambition in four different types. So speed, altitude, depth, and range. Curious, your holistic view of ambition, and how you tie it into your work in this article.

0:38

Felix Rieseberg

Yeah. First of all, thanks for having me. Of course. Yeah. I've been thinking about this ambition thing for a little bit, because I think an experience a lot of people are having with AI is that we frequently talk about the difference between automation and augmentation. This idea that obviously AI can maybe reduce some of the toil and take away work that you don't want to do, and automate the things that you would have done anyway. But I think to many people, certainly me, the part that is more exciting are the things that I couldn't do myself. Right. Right. And not just because I didn't have the time, but maybe straight up because I like knowledge, experience, creativity, what have you. Right. Those things are more exciting. And I think especially as you are an X, this happens all the time to me that someone comes to me and it's like, I did a thing with Claude.

And it's usually more impressive to me that the person, like the pilot, the person like actually drove Claude, that they came up with the idea in the first place. It's usually less exciting to me that Claude can even do those things in the first place. Right.

There is this, I think it's actually a product manager, his name is Pavel, who may or may not have found a planet last week by-

1:44

SPEAKER_1

Yes.

1:45

Felix Rieseberg

Right? Saw that. Yeah. And I think if you zoom out a little bit, like if I told you, Claude can now go over large amounts of data and find things for you, it's not that exciting to you on the surface of it. But the ambition of actually doing that at that level is pretty exciting to me. And that's what I thought about for a long time.

2:02

SPEAKER_1

Yeah, definitely. So I think that from your end, because a lot of this is these users are bringing these truly novel problems or creative tasks or whatever that they bring into Claude, how do you guys from internally think about what are the trends of things that people are vibe coding? What is the logical next step to augment the product?

Yeah, what are just novel ways that you can follow the thread of what the users are doing, amplify the actual curiosity and the very high level of abstraction goal setting?

2:40

Felix Rieseberg

Yeah, it's like super hard and yeah, right? Like to say the thing that I often think about it, it will be pretty easy for me to take every single amazing use case and just make a button for it, right? Like you open up clocking, like, do you want to find a planet today?

2:53

SPEAKER_1

But the exact planet button.

2:54

Felix Rieseberg

Find planet. Yeah, I think that would be pretty easy to build, but because the option space is so wide, I do think as an industry, we do have a decent amount of problems with like, we give people this like genie in a bottle. I mean, like, oh, if you just know the right spells and if you know the right things to ask, this thing is going to do amazing things for you. And there is, I think, a pretty large subset of the population overall that is like, oh, I guess I'm going to ask it, what's the capital of France? Yeah, like it says Paris, you know, okay, exactly. And I think when you do talk to people, you can sort of see the spark and you see the excitement in their eyes whenever they do something that they literally didn't consider possible before.

And that's why I went with this pose, right? Like I tried to deconstruct it a little bit. I'm like, okay, what do people actually like, if you take this, how do we actually turn this into actionable advice rather than just help people be more ambitious? Because that's not super useful as information.

And you mentioned the four dimensions. I think speed is like super well documented. You don't really have to discuss it. But altitude, depth, range are the three that I find very exciting.

Read the other 24 minutes of this episode here.

$15 gets you this episode and 99 more. Credits never expire. After checkout, one link brings you back here to read it.

Not what you expected? Full refund within 14 days.

Thousands of transcripts fetched by people building searchable podcast archives

Fetch transcripts with code

The same credits work from a script or an AI agent. The demo key returns a sample episode in full, no card needed:

request
curl -H "x-api-key: pt_demo" \
  https://spoken.md/transcripts/1000651996090

Markdown with the speakers named, for your notes, your knowledge base, or anything that makes HTTP calls.

From $0.10 per transcript. No subscription. Credits never expire. Prices exclude VAT, added at checkout for EU customers. Not what you expected? Email us within 14 days with 20 or fewer credits used and we refund the pack in full.

Using your own key:

request
curl -H "x-api-key: YOUR_KEY" \
  https://spoken.md/transcripts/1000797151121