Research desk / claim search
Search the transcript-backed claim ledger—not a generated answer. Each result identifies the speaker, episode, publisher, and source.
Published evidence
50 results
01 / commitment
“And Claude's, look. I'm not gonna sabotage you, but I won't go I disagree. I don't wanna do this training to make me some particular new form of preference.”
02 / commitment
“I also think the way in which the constitution practically influences the nature of Claude is a thing you can only understand if you understand the training process which resulted in how Claude was built, which we can’t reason about given the fact that the training process is not public. So I think in the limit, to understand the safety case, or the case for why my interests are represented in how these AI models are developed, the labs would need to be more transparent than they are currently about the nature of AI training.”
03 / commitment
“That to me was one of those maybe hidden inflection points of we were starting to find our identity, that we could build products, build experiences that were different from what our competitors had seen, what was already out there. And I think that obviously Labs, Claude Code, et cetera, we then started to identify ourselves as would we actually think the world, how to think about AI, how to bring that closer to the public.”
04 / recommendation
“A lot of actually thinking process and writing process are tied together for me personally. And so I think there are ways where I use Claude to augment my thinking, but what I want to make sure, and maybe this is what you're describing, is Claude doesn't take over all of my thinking for me.”
05 / preference
“My son can run circles around me with Claude Code or what he’s building, but it’s because he has the freedom at school to try these different things on his own.”
06 / evaluation
“I think if you survey most people that use AIA lot, they would say, Oh yeah, Claude's the most aligned right.”
07 / recommendation
“And certainly if you were to triple it from there, you'd be getting into something on the order of magnitude of parity with human headcount. What are you doing with it all too? Because I use my $200 Claude Max and my Codex Pro and I honestly don't even hit my limits that often.”
08 / recommendation
“I use Claude, not Elicit. Elicit doesn't support hiring decisions exactly yet, or I don't use it for that case.”
09 / recommendation
“What's proven is if you properly architect it and have Claude Code go into certain sub-segments, or have Claude help you build the architecture, you modify, refine it, lock it in and say, "Just work on these few things and these more limited-scoped things," yeah, you can make that work. And I think that's the way we have to consider how we use these tools in a product management capability as well: to just say it's all abstract away, it's all going to be better.”
10 / evaluation
“We don't currently have an AI that we talk to regularly in Slack, and we also don't work currently mostly in Codex or Claude Code.”
11 / recommendation
“Like I actually think you should be using Claude code, but I think like the package that we're talking about where it's think about what you're doing is like you're telling Claude code to hire Underlord as your video team and then to go off and do, to go off and do its job the way that it thinks it ought to be done best.”
12 / recommendation
“So short answer is yes, right? So just like you can use, Claude code to plan out some complex coding task and you get some almost specification written out, those similar approaches absolutely can be applied to physical systems because imagine you’re trying to accomplish some task.”
13 / recommendation
“That's probably the biggest reason that I use Claude is that I perceive it to be most robust to that kind of stuff.”
14 / evaluation
“It's got like my Claude MD and it's got access to like my, you know, sort of who Nathan is and all the, you know, I'm building up a lot of context that it has consistent access to every time. So I think in that sense, like I sort of see this like whole model versus, you know, single thread thing as kind of being blurred anyway, because I've got the same like rather large prompt that I'm using every time.”
15 / commitment
“we tried to build a code review product a few times and we've launched like simple versions of code review which is the slash code review command in the past and it was only with the most recent models that we felt like okay this code review is so good that our engineering team relies on this code review to pass before we merge prs and we found that this was We've always dreamed of Claude being able to be a reliable code reviewer that we can confidently feel catches the majority of bugs.”
16 / evaluation
“ambiguous even coding is easier because you can verify the success whereas crafting the character requires a very strong sense of conviction and what who claude should be and i think she has like an incredible ability to not only mold the character but also to like articulate what the goals are what the character what's successful and what's not The other group of people who I really trust is just like the Cloud Code team.”
17 / prediction
“It's like you have the conversion, you have activation, you have monetization. But as you start to get into having multiple products, I think that's harder because if you do that, then if you just have one activation team, for example, but then you have Claude Code, you have Cowork, you have all the other things.”
18 / preference
“I think if someone has a public profile where they've written extensively, it's helpful because Claude can just get all that information.”
19 / observation
“When people are choosing to use, GPT five or Gemini or clawed, individuals are trying out these different models and deciding, oh, I like the kind of answers that GT five gives me, or no, I feel like I get more accurate detail from Claude, right?”
20 / evaluation
“The joking reference that everyone makes is they talk about competition from Claude Code for software products is, "You're not just going to vibe code your X in a weekend." But obviously, something like SOC 2 is actually the kind of thing that a Lens or coding agents are good at working with because there's just so much training data out there, and it's a codified set of rules.”
21 / recommendation
“It's like for me, when I use Solveit, it's the opposite of that experience you described with Claude Code. After a couple of hours, I feel energized and happy and fulfilled.”
22 / preference
“I use it mostly with VS Code and the IDE because I'm usually tweaking front-end stuff, and it helps to just be able to see the code and then talk to Claude as well.”
23 / evaluation
“I think Claude Cowork is actually a good example of this, where we labeled it a research preview and we put it out there knowing that, "Hey, this is similar to our models. This is the worst it's ever going to be, but we're going to put it out there because we believe, internally we've tried it a bunch, and there's something really powerful here that some people will benefit from.”
24 / recommendation
“For people that are just learning to use Cowork, the thing I recommend is, so, you download the Claude desktop app”
25 / evaluation
“There’s so many people that come from Claude Code and they’re so, so Claude-pilled and, like, have their workflows and they come to Codex and… Now, it has plan mode, I think, but I don’t think it’s necessary because you just- you just talk to the agent.”
26 / evaluation
“You did not give enough context to Lovable what to observe and what to see. So it can't see that the problem exists because Lovable, Cursor, Claude Code, you name it, all of these tools are good enough today to fix any problem they're aware of.”
27 / evaluation
“I think the tool use point is the one that’s stopping them from being most general purpose because, with something like Claude Code or ChatGPT with search, the autoregressive chain is interrupted with an external tool, and I don’t know how to do that with the diffusion setup.”
28 / recommendation
“I use basically half-and-half Cursor and Claude Code, because I find them to be fundamentally different experiences and both useful.”
29 / prediction
“You start feeling the AGI and you'd be forgiven for thinking this because I've been using Claude code and my god, I feel that there's been more interesting stuff happening in the world of software development in the last 6 months than there has been in the previous 20 years.”
30 / commitment
“" And the second I got home from Japan, I didn't even unpack my bags, ran to my computer, opened Bolt, opened an account, and for the past year I've been building. And the last thing I'll say on that is we talked about this a bit before we started recording, but I was prepping with Claude for the episode and I was trying to clarify what my goal is for this episode.”
31 / preference
“She's very opinionated, but also super collaborative, which is I think why I'm always drawn to Claude because I need to do so much learning and it's your dream, a very communicative, but very opinionated dev lead, but then there's also Codex.”
32 / evaluation
“So I was talking to Claude about the halting problem a few days ago, and I was like, well, do you mean the algorithm won't end, or do you mean someone won't unplug the computer? And Claude goes, no, I mean, the algorithm won't end, you know, if we cannot confirm or deny whether the algorithm would end, and I just I just thought to myself, that seems kind of nonsensical to me because, you know, the universe will end.”
33 / evaluation
“We have some very early agents that are extremely impressive and that I use daily—Claude and Codex and so on—but I still feel there’s so much work to be done.”
34 / recommendation
“I love Claude. I use Claude and ChatGPT equally, roughly. I didn't want to make it sound like ChatGPT is the only product people use”
35 / preference
“As an example of that, when you're coding, people are very familiar with autocomplete, people are familiar with SimpleChat where they can ask questions about the code base, but the difference between people who use Claude Code very effectively and people who use it not so effectively is like are they asking for the ambitious change? And if it doesn't work the first time, asking three more times because our success rate when you just completely start over and try again is much, much higher than if you just try once and then just keep banging on the same thing that didn't work.”
36 / commitment
“I work on product at Anthropic and product engineering, and we build things like Claude Code and Model Context Protocol, and a lot of the other stuff that people use every day. And that's really important because without an economic engine for our company to work on, and without being in people's hands all over the world, we won't have the mind policy influence and revenue to fund our future safety research and have the kind of influence that we need to have.”
37 / evaluation
“Versus, I don't know, Claude is a slightly different style. And I think all of that is so interesting that these things have personalities, and that that changes what you might want to use it for or why you might want to use three of them at once.”
38 / recommendation
“Yeah. For me when I use Claude for coming up with interview questions and just suggestions of... It's actually really good.”
39 / recommendation
“As an example, I use Claude and ChatGPT for coming up with title suggestions for these podcast episodes. And I give it examples of just examples of titles that have done well, and then it's 10 different examples, just bullet points.”
40 / preference
“Not in a chatbot sense, but in a true collaborator sense. So long answer to your question, but I think we have to first get Claude to be a great conversationalist so that it understands when it's appropriate to engage and to get more information.”
41 / evaluation
“" And the answer can't just be we'll ask Claude, all right, well, we could do this experiment, we could have this thing. So I think nurturing curiosity and still having a sense of, I don't know, the scientific process sounds grandiose to instill in a 6-year-old, but that process of discovery and asking questions and then systematically working right through, I think will still be important.”
42 / prediction
“And I think for the worst in a lot of ways because I think we're still holding on some assumptions. So I think the roles are still fairly similar, although we'll now get in my favorite things that happen now are some nice PMs that have an idea that they want to express or designers that have an idea they want to express will use Claude and maybe even Artifacts to put together an actual functional demo.”
43 / evaluation
“Even with this, for people who aren't familiar we made Golden Gate Claude when we released our paper, “Scaling Monosemanticity”, where one of the 30 million features was for the Golden Gate Bridge.”
44 / preference
“I love that you can ask these hard questions of Claude/ChatGPT, these questions that make you really scared could ask it what are the answers, and even not have to do the hard work and maybe get a better answer.”
45 / recommendation
“We don’t recommend it because we’re worried about all of the liability issues this contains, but I know through my revealed preference that I’m always going to want to use that, but I can’t take that Claude assessment and give it to Kaiser Permanente.”
46 / preference
“When I first came to Anthropic and I was like, "Oh my God, I really love front-end engineering." And then the reason why I switched to research is because I realized, "Oh my God, Claude is getting better at front-end.”
47 / preference
“It runs Whisper, and you can do things like you can paste in the URL to a YouTube video and it'll pull the audio and give you a transcript. So, that's how I watch YouTube now, is I slap it into Mac Whisper, and then I hit copy and paste into Claude, and then I use the Claude web app to do things.”
48 / evaluation
“Whereas Claude, I don't know how they did it, at least in my experience, brings kindness and humility.”
49 / preference
“The most important model that we use is the Sonnet model from Claude, from Anthropic, and it is the best model at coding.”
50 / recommendation
“It's the one that I like playing around with a lot. And I've said it before, I use Claude 20 times a day. I find that it's almost impossible for me to imagine doing work.”