How to get the most out of AI with Nick Lichtenberg

Summary

---

Podcast Summary: Fixable – “How to Get the Most Out of AI with Nick Lichtenberg”

Podcast: Fixable
Host(s): Anne Morriss, Frances Frei
Guest: Nick Lichtenberg, Editor at Fortune
Date: August 3, 2026

---

Main Theme & Purpose

This episode explores the real-world impact of generative AI in the workplace, zeroing in on how journalists—and specifically Nick Lichtenberg—are leveraging AI to boost productivity, efficiency, and the breadth of their work. The discussion candidly addresses both the hype and skepticism surrounding AI, aiming to ground the conversation in practical examples and actionable advice.

---

Key Discussion Points & Insights

1. How AI Is Changing Journalism

• Nick’s background in business journalism: Started pre-Great Recession, lived through the seismic shift when Craigslist displaced classified ad revenue.
  • “Now the internet age is giving way to the AI age, and we’re still wrestling with the internet problem, and AI is giving us a whole new business model problem to chew over.” (05:53, Nick)
• AI as a tool for coping with information overload:
  • “AI helps me digest maybe like from three or four years ago, a whole day's worth of content in an hour. So I look at it as like a compressor of information in a good way.” (06:36, Nick)
• Quantity vs. Quality in Content Production:
  • AI enables the production of more content at lower cost, but also presents new challenges for maintaining standards.

2. The Human Element and Guardrails

• AI is not replacing humans but augmenting them:
  • “It can’t think for itself and it really cannot take our jobs, is my firm belief... it can do it for me quickly and easily, but it can’t ask itself the right question, and that’s my job.” (01:08, Nick)
  • “My metric is let’s tip the excitement a little higher than the fear today.” (04:51, Anne)
• Guardrails: The Need for Fact-Checking and Editorial Judgment:
  • Human oversight is essential—AI is agnostic about source credibility, so journalists must verify every claim.
  • “It matters a lot to me that I write stuff that I’m proud of. Even the quick stuff where I am using AI to crunch a lot of notes and documents together, the 50% me...is that the angle you dive into the story, the personality...that’s the 50% that comes from me.” (28:15, Nick)
  • “I have to keep pushing my brain and my team and my AI coworkers now to really say what we’re trying to say, to really pull something off.” (31:56, Nick)

3. Practical Takeaways: How Nick Uses AI

• Workflow Example:
  • Input: Dozens of analyst notes or complex documents.
  • AI summarizes, finds common themes, provides citations.
  • Human double-checks all citations and sources, corrects errors, and adds perspective.
  • “I asked it as a follow up, cite all of your sources here. And it did that. So I was able to then kind of painstakingly verify all of the quotes and the stats that it compiled for me.” (21:50, Nick)
• Risks of Hallucination:
  • AI sometimes generates plausible-sounding but false information:
    • “At the very end of the article, there was a quote that looked wrong and there was no citation for it...and it said, you got me. I just fabricated that one because it fit the rest of the article really well.” (23:03, Nick)
• Best Practice:
  • Structure prompts to demand sources and authenticity.
  • “The human judgment part of it has been super important for me. ... Building that into the prompting… creating rules for the output… constantly pressure testing those.” (23:28, Nick)
  • “Include citations for all the facts and the quotes that you found here.” (24:48, Nick)

4. Broader Reflections on AI & Work

• Not a zero-sum game:
  • AI frees up humans for higher-level creative and strategic work.
  • “You create more time to have conversations like this one, because you have a robot assistant who is compiling spreadsheets for you and compiling tables of content for you and outlines for you and all these things that you just don’t have to do anymore.” (39:08, Nick)
• Learning from the Extremes:
  • Nick as a productivity outlier—and what others can learn from experimentation at the boundaries.
  • “I personally find that great learning happens at the extremes...one of the reasons, Nick, that you’re a lightning rod...is because you are learning at the extreme.” (15:51, Frances)
• Concerns about “Average” Output:
  • Frances fears AI might encourage mediocrity. The ideal is to use AI to make us “outliers,” not average.
  • “One of my greatest fears about AI is we lose our distinctiveness, that it all becomes average...But if it helps us become even more outliers than we were before, fabulous.” (04:33, Frances)

5. Skills for the Next Generation

• Machine Learning Literacy as Critical Skill:
  • “Machine learning literacy might replace media literacy as the big challenge going forward.” (12:38, Nick)
• Advice to Young Professionals:
  • Read broadly, compare multiple sources, and invest in developing your own judgment.
  • “There was advice not to read one newspaper but to read four newspapers and then compare how they chose to assemble the facts.” (36:45, Nick)

6. Predictions and Pep Talks

• The Pace of Change:
  • Nick is “bearish”—predicts slow, incremental changes rather than disruptive revolutions.
  • “I don’t think we’ll see mass disruption of jobs...AI will be like electricity; some think it’s happening super fast...but it’s just not that simple to rewire everything.” (37:41, Nick)
  • Frances is more optimistic ("it’s going to go much faster"), citing fast learning by extremes and high-performers. (38:26, Frances)
• Experimentation Mindset:
  • Everyone should try AI as an "assistant," experiment, and be open to reimagining their jobs.
  • “Think of it as your assistant that’s doing all the stuff that you don’t want to do...you collect the materials, you give them to your assistant, and they do it for you, and then you proofread them.” (39:08, Nick)

---

Notable Quotes & Memorable Moments

• On the Fear of AI:
  • “If it was called mass information processing computer plugin, people would feel differently...something about the phrase artificial intelligence is making this more provocative and harder to adapt to than it needs to be.” (11:15, Anne paraphrasing Nick)
• On AI's Limits and Human Judgment:
  • “AI...can’t ask itself the right question and that’s my job.” (01:08, Nick)
• On the Role of Feedback:
  • “When you do tell them [AI tools], this is a good example, this is a bad example, they will change their output to match the examples you give them.” (27:04, Nick)
  • “Make sure you give them feedback, they’ll hallucinate that one thing and you’ll be like, oh, I don’t know about this assistant.” (40:33, Frances & Nick)
• The “Sandbox” Approach:
  • “I hope that a lot of people are free to experiment and reimagine their job...most people will be doing more interesting work.” (32:51, Nick)

---

Timestamps for Important Segments

• 01:08 – Nick’s stance on AI’s inability to replace human jobs
• 05:13 – Context: What problems does AI solve in journalism?
• 06:36 – AI as a tool for information compression
• 13:43 – Speed and nature of the newsroom before/after AI
• 17:10 – Importance of fact-checking, risks of misinformation
• 21:50 – Practical example: Using AI to analyze market analyst notes
• 23:03 – The reality of AI quote “hallucinations”
• 27:04 – On giving AI feedback and improving output
• 28:15 – Nick’s “50% human” articles & pride in authorship
• 31:56 – Concerns about AI “doing our thinking for us”
• 36:45 – Advice for developing news/media literacy in the AI era
• 37:41 – Nick’s prediction for AI’s pace of workplace change
• 39:08 – The “assistant” metaphor and encouragement to experiment

---

Tone & Style

The conversation is candid, insightful, and practical, with a balance between optimism and healthy skepticism. Nick is pragmatic about AI's capabilities, stressing the core importance of human judgment, care, and creativity. Frances and Anne ground the discussion in both their excitement and their existential worries, ultimately providing a calming, realistic lens on AI’s actual role in creative and knowledge work.

---

Wrap Up

This episode of Fixable moves beyond the hype around AI, highlighting the lived – and still evolving – experience of one journalist at the frontier. Listeners are encouraged to approach AI not as a threat, but as a powerful, unpredictable tool—one that amplifies both the opportunities and responsibilities of human work. Nick’s story stands as both a case study and a call to experiment, learn, and maintain the highest standards, even (or especially) as our tools change.

---

Transcript

[Capital One Sponsor]

This episode is brought to you by Capital One. Capital One's tech team isn't just talking about multi agentic AI. They already deployed one. It's called Chat Concierge and it's simplifying car shopping using self reflection and layered reasoning with live API checks. It doesn't just help buyers find a car they love, it helps schedule a test drive, get pre approved for financing and estimate trade in value. Advanced, intuitive and deployed. That's how they stack the that's technology at Capital One.

[Progressive Sponsor]

You're listening to this podcast, so I know you've got a curious mind. Here's a helpful fact you might not know yet. Drivers who switch and save with Progressive save over $900 on average. Pop over to progressive.com, answer some questions and you'll get a quick quote with discounts that are easy to come by. In fact, 99% of their auto customers earn at least one discount. Visit progressive.com and see if you can enjoy a little cash back. Progressive Casualty Insurance Company and affiliates national average 12 month savings by $946 by new customers surveyed who saved with Progressive between June 2024 and May 2025. Potential savings will vary.

[Nick Lichtenberg]

I always have to grab the reins at some point and take over because it can't think for itself and it really cannot take our jobs is my firm belief. And I'm using it for a very specific purpose, which is that I've got all this information that I need to process and turn into journalism and this can do it for me really quickly and easily. But it can't ask itself the right question and that's my job.

[Anne Morris]

This is fixable. A show where we take the problems at work that feel unfixable and prove that they're not. I'm Anne Morris.

[Frances Fry]

And I'm Frances Fry.

[Anne Morris]

Today on the show we're excited to dive into a real life AI use case that's causing a stir in and beyond the world of journalism. Our guest today is Nick Lichtenberg. He's a veteran business journalist who made headlines of his own earlier this year with a much talked about Wall Street Journal profile. Nick is an editor at Fortune, where He's published over 600 articles so far this year with a big lift from AI. He uses the technology to research, edit and write many of his stories. AI assisted articles are a big part of Fortune's editorial strategy. At one point they accounted for about 20% of a site's web traffic and many of those were Nick's stories. And as you can probably Guess. People in and out of media have had a very strong reaction to Fortune. And Nick's approach is why we wanted to bring him on the show. Our curiosity was piqued. We're still hearing more from leaders about AI's potential than its real life transformative impact. But Nick is the real deal in the sense that he's actually doing the thing that everyone says AI can do. He's insanely productive. He's publishing seven articles a day. But he isn't letting AI drive the proverbial car. He said that AI is like a sports car you can crash if you're not careful. You gotta be a Formula one driver. So we're gonna get an AI precision driving course from Nick today and hear how he handles this very powerful car. Including what guardrails he's put in place to avoid an all out collision.

[Frances Fry]

I have so many questions including what are his hopes and dreams for the field that he's in. But mostly I'm gonna know how he did it because I am. I am more productive with AI. A little bit more. What it has given me is I can do a set of things I used to not be able to do before. So it's allowed me. It's not really productivity. It's like going into the unknown. I don't even know what it means to be this much more productive. It's really quite astonishing. And so I have so many questions.

[Anne Morris]

Yeah. And I want to get into all of that. I feel like you're an early adopter like Nick. Maybe not all the way to Nick, but you're directionally a lot closer than I am. I'm trailing behind both of you probably with a little bit of skepticism and probably with some fear of losing something precious in this revolution in myself and in the world. I'm hoping today to be outnumbered in the best sense. I want to feel some of your excitement and infectious wonder about this technology.

[Frances Fry]

You know, one of my greatest fears about AI is we lose our distinctiveness, that it all becomes average. Like this is like the great converge to the mean. That's my fear. But if it helps us become even more outliers than we were before, Fabulous. So that's where I want to go with it.

[Anne Morris]

All right. My metric is let's tip the excitement a little higher than the fear today. And then I think that's success.

[Frances Fry]

Awesome.

[Anne Morris]

All right. Let's bring Nick in. Nick Lichtenberg, welcome to Fixable.

[Nick Lichtenberg]

Thank you so much. Good to be here.

[Anne Morris]

I want to start in a very practical place. I promise the Questions will get more interesting, but I think this context will be helpful for people to just absorb your story. So from your perspective, what problems are you solving with AI for yourself and also for your industry?

[Nick Lichtenberg]

That's a great question, and it's a simple but complex one. My whole career in journalism started around 2005 and 2006, a little bit before the Great Recession. That was after the first real crisis in the business model of journalism with Craigslist.

[Anne Morris]

Oh, yeah.

[Nick Lichtenberg]

It's well known to me, but not maybe to many people, that most newspapers relied on classified ads for a huge part of their revenue, and that Craigslist almost immediately wiped that out for everyone. And now the Internet age is giving way to the AI age, and we're still wrestling with the Internet problem, and AI is giving us a whole new business model problem to chew over. Because I think for as many problems as AI solves, it creates new ones. That's the crux of the issue, maybe why I'm on talking to you today. But it solves the need for more content at scale in the content space.

[Anne Morris]

So the quantity of content at the same cost is the major structural problem it solves.

[Nick Lichtenberg]

I think that's one half of it. And then the other half is the amount of content that journalists have to digest before they produce a story. AI helps me digest maybe like from three or four years ago, a whole day's worth of content in an hour. So I look at it as like a compressor of information in a good way. So we hear garbage in, garbage out is a problem using AI. But if you have good inputs, you can have good outputs is what I believe.

[Anne Morris]

Yeah. You told the Wall Street Journal, I know that this won't be seen as some people's idea of journalism, but I've always seen myself as a kind of a nobody from nowhere who strapped and clawed to get into this industry and find solutions to problems. What's the connection between those two ideas for you? AI isn't real journalism, and you're an outsider who had to fight his way into this arena.

[Nick Lichtenberg]

That's great thing to zero in on. My parents didn't love it when they saw that quote in print. They said, you're a somebody from somewhere. And I said, yeah, well, I think what I was trying to conjure with that is that there are many different types of journalism, and some of them have been using AI for a long time. They have really deep histories. And that's sort of how I came in and kind of through the side door, so to speak. I worked for Several trade publications, several really business focused publications. So when I was at Bloomberg, they had something called Bloomberg Automation and they had a team of machine learning experts and it was super helpful, very disruptive for earnings reports. And before Bloomberg Automation, it was someone's job to read the earnings report as soon as it dropped and race to find the right number and then immediately flash that headline to the terminal. And by the time I got there, this is about five years in, this was 2018 or so, Bloomberg Automation was finding that number itself in a millisecond and publishing. And your job as the human was to back read Bloomberg Automation and make sure the number was right. And sometimes it was off by a penny. And that mattered a lot to the stock market. I met a Bloomberg editor actually at a media happy hour shortly after the Journal piece came out. And we didn't know each other beforehand. And his eyes kind of widened when he recognized me from the Journal article. And he said, oh my God, how are you doing? And then we talked about that for how I was doing with this weird thing that happened to me, becoming a main character on the Internet, so to speak. And then he said, you know, I work at Bloomberg now. You used to work at Bloomberg. Like what we do is not really writing. It's using tools to get information out for a financial services focused audience. And that's been the case for a long time. And so the work that I do for Fortune and then getting profiled and revealing this to the world, it was like a collision of those two worlds to me. And Fortune has one foot in both worlds. And I don't think I fully appreciated that before at all. Kind of kicked off for me.

[Anne Morris]

Well, let me ask you the same question. How are you doing as like the poster child for this revolution and for everybody's big feelings about this technology?

[Nick Lichtenberg]

Yeah. I was really puzzled at first when I became a poster child. I actually just got a request in my LinkedIn from a journalism professor to talk to one of his classes. And he called me a pin cushion for the industry. And it hurts my feelings sometimes, but I also am more interested in what's happening. I write about other people for a living, and now people are writing about me a little bit and you worry about what they're going to conclude about you. And it's a lesson in being the subject of journalism as well as a proprietor of it. So it's fun, but stressful.

[Frances Fry]

Nick, the hardest thing I have found to deal with, if you're profiled in some way, it's not even the scrutiny it's the sympathy I find that is

[Anne Morris]

the emotion that my wife cannot tolerate.

[Frances Fry]

I just found it close to unbearable.

[Anne Morris]

I'm fine with it. Bring on the compassion. You said at one point. Sometimes I think if it wasn't called AI, if it was called mass Information processing computer plugin, then people would feel differently. I 100% agree with that. I do think that there's something about the phrase artificial intelligence that is making this more provocative and harder to adapt to than it needs to be. Will you say more about that?

[Nick Lichtenberg]

Yeah. One cool thing that's happened as a result of this profile of me is that more people know about me and are asking me to come and talk about my experience. And there is a lot of fear and really good discussion and some pointed questions. But when I say, like, this stuff is not intelligent, it's just a computer tool and I always have to grab the reins at some point and take over because it can't think for itself and it really cannot take our jobs, is my firm belief. And I'm using it for a very specific purpose, which is that I've got all this information that I need to process and turn into journalism. And this can do it for me really quickly and easily, but it can't ask itself the right question. And that's my job. And so it's kind of amazing when you use it the right way and then it's super frustrating and even bizarre when you use it the wrong way.

[Anne Morris]

And.

[Nick Lichtenberg]

And I've been saying that machine learning literacy might replace media. Literacy is the big challenge going forward because we're going to need to understand that AI tools really can only digest the information we feed them. And if you have a 10,000 word manuscript, that becomes the Bible to the AI machine that you're working with. And if that manuscript is flawed in its own design, you'll get a flawed response.

[Anne Morris]

I do think that's one of the things that's really helpful about, you know, even for people not in journalism or media. The use case of Nick Lichtenberg using these tools and kind of transforming his own experience of his work. This conversation comes down to the very practical. And it's not as if to your job's point, you were deciding between, oh, do I hire three people to help me write seven stories a day.

[Frances Fry]

Yeah.

[Anne Morris]

Or do I work harder to try to create something like this was the only thing. There wasn't a human substitute to what you're doing with these tools.

[Nick Lichtenberg]

Yeah, I would say so. And the workload doesn't actually feel different to Me, I've had very stressful, stressful business journalism jobs my whole career. Like when I was at Bloomberg on the breaking news desk, that's like a fire hose of information. And they measured our work in how many seconds faster we were than our competition at Reuters. But then also the number of screens we could fill up with headlines on a busy earnings day. And I would sometimes have 12 screens full of content, which was, I think, 120 unique headlines that were being sent in real time to the stock market. So this work is maybe slower than that work.

[Frances Fry]

This is Nick being Nick.

[Nick Lichtenberg]

Yeah, I think this is me being me. I think by necessity I had to develop the skill of absorbing information. And AI is just another tool to do that. But you have to use it right. And I wouldn't still be here talking to you guys if I hadn't figured out a way to build stories with this tool that people actually want to read. Like, it's not AI slop to me because it is actually landing with people and they find it interesting and informative. And I would point to some of our reporting on the petrodollar system, which is all about the Iran war and the Strait of Hormuz and how the dollar system, it works with oil being acquired in US dollars and then being reinvested back into treasury bills. And AI helped me pull together information that was already all out there, but not connected in that way. And then I saw really smart analysts, much smarter than me, getting published in great newspapers I love, like the Financial Times, talking about the petrodollar news cycle. And I was a part of that, and AI helped me be a part of that. So why is that a bad thing? I still don't really get that.

[Anne Morris]

Yeah, we're going to get into the right way and wrong way to use these tools. For instance, any more big questions for Nick before we get to tactical?

[Frances Fry]

It's an observation which is I personally find that great learning happens at the extremes. So if you ever watch me teaching a class, I'm ushering students to the extreme parts of a conversation in order to really figure out the limits of what we're talking about. I think one of the reasons, Nick, that you're lightning rod, I think, is the word that was used earlier, is because you are learning at the extreme. Let's be honest, 600 articles this year, seven articles in a day. Those numbers, number two in productivity is nowhere near. And so that, to me, might be part of one, part of why people are so fascinated. So learning at the extremes I love. But if I look at what you're doing. AI is making you broader. Like, it's making you broader. And in some ways I could imagine calmer. And not the stress part of it, but the calmer in the sense that, like when you. You probably now think you know about the petrodollar system in a way that you might not have if you were just going to the library and checking out books. Because how do you know you have all of them? How do you know? So I actually think there is this broadness and calmness to how you talk about it. That's kind of cool.

[Nick Lichtenberg]

Well, thank you. I would agree with that. But I think the thing that is scary to me and the right and the wrong way to use this is that just the fact checking involved. And I'm not claiming to be like at the New Yorker where they have a famous fact checking department that actually tracks people down to their house to make sure they are who they say they are.

[Anne Morris]

May those people always have jobs forever. Like, it's a beautiful public service.

[Nick Lichtenberg]

Yeah. But it's important to me that every single claim that the AI tools that I use bring to me that I'm tracing those down to the original source. So this petrodollar system, like, what is the actual reporting that first revealed this? It was a Bloomberg FOIA request about 10 years ago, actually. And then what is all the supporting evidence and how reliable are all those sources that you get? Because the AI is agnostic about reading the Internet. It'll read the Internet and bring back whatever fits the prompt.

[Anne Morris]

The Reddit rantings of our 18 year old son or the Reddit are treated equally. Yeah.

[Nick Lichtenberg]

And someone said to me, like, when are people going to start planting intentionally fake things on the Internet just to confuse the AI tools? Like then, then these tools become way less useful. And maybe that's already started happening.

[Anne Morris]

This episode is powered by AT&T Business. I was thinking recently about those early days of building something of your own. It's not just the little things. You're building the whole plane as you fly it. Think of those mornings you might find yourself sitting in a crowded coffee shop or the back of a library, hunched over a laptop and just hoping the public wi fi would hold long enough to upload a pitch. It's a stressful way to start a day and an even harder way to build a legacy. You're working from wherever you can, piecing things together, hoping everything holds. And it's funny, connectivity is one of those things you don't really think about until it becomes a problem, and when it does, it can throw everything off. The last thing you want is to be worrying about whether things are going to work when you need them to. That's why AT&T business is a great provider for small business owners. It's built to work so you can stay focused on what you're building. Powered by AT&T Business Built to Work get today at business.att.com support for today's episode comes from Square. Think about the last time a local business made your day a little better. Maybe it was a quick checkout, a friendly hello, or someone remembering exactly what you like. For me, it's that perfect cappuccino from La Colombe. Even on a busy morning, Square helps the people running these shops to keep things moving so they can focus on creating moments that matter. As a business, you can track sales, manage inventory and access reports in real time, whether you're in the shop on the go or running things all on your own. With Square, you get all the tools to run your business with none of the contracts or complexity. And why wait? Right now you can get up to $200 off square [e-mail] go fixable. That's sq u a r e.com geo fixable. Run your business smarter with Square and get started today.

[Capital One Sponsor]

This episode is brought to you by Capital One. Capital One's tech team isn't just talking about multi agentic AI. They they already deployed one. It's called Chat concierge and it's simplifying car shopping using self reflection and layered reasoning with live API checks. It doesn't just help buyers find a car they love, it helps schedule a test drive, get pre approved for financing and estimate trade in value. Advanced, intuitive and deployed. That's how they stack. That's technology at Capital One.

[Anne Morris]

Let's get practical. Let's get into the right and wrong way to use this. So in this conversation on the spectrum of early to late adoption, I think you're farthest to the left on early adoption. Francis is next and I'm. I'm further away now that we're living in the Nick Lichtenberg era of journalism. Like without giving away your trade secrets like so how do you think about constraints and guardrails when it comes to your use of AI?

[Nick Lichtenberg]

Yes, this just happened to me this morning so I put together a piece that I could not to Francis's point, it broadened me in a way that I would have struggled to do in the time allotted. Earlier in my career, Micron Technology had big earnings earlier this week. And many analysts think that it single handedly stopped the global tech sell off that was really hurting financial markets and all of our 401ks. And we found about a dozen different notes from Wall street analysts. And I asked AI to tell me what the common themes were and what the conclusions were. And it helped me put together a piece that I'm really proud of. And then I asked it as a follow up, cite all of your sources here. And it did that. So I was able to then kind of painstakingly verify all of the, the quotes and the stats that it compiled for me. And then at the very end of the article, there was a quote that looked wrong and there was no citation for it. And I said, where did this come from? And it said, you got me. I just fabricated that one because it fit the rest of the article really well.

[Anne Morris]

I'm sure this is going to get fixed. But that is such a crazy part of interacting with these tools. I'm like, that doesn't seem like a real person. And they're like, oh yeah, my bad, my bad.

[Nick Lichtenberg]

Yeah. And I think like no matter how many times you tell them to act a certain way, they will not reliably do that.

[Anne Morris]

Yeah.

[Nick Lichtenberg]

But then they will do 90% of what you want.

[Anne Morris]

So indulge us with a little bit of like play by play. So how do you, how are you using these tools in a typical article?

[Nick Lichtenberg]

So I also was on a panel recently and I said, I really believe in liberal arts education. I really believe in hard skills, because those are the things you really need to be effective here. And that leads me to say the skills I think we really need are a rock solid understanding of what editing writing looks like and what solid information looks like. And so that will help you verify everything you're seeing. Like the obviously fabricated quote that doesn't match up with all the real quotes that it's sitting next to. So I think the human judgment part of it has been super important for me and building that into the prompting. And I've built a few agents too. Although I wouldn't claim to be an agent wizard. Like, if you guys think I'm an early adopter, there's agent programmers much better than me. But it's essentially creating rules for the output that you want and then constantly pressure testing those and checking your work and saying like, do we make sure that we got all of this buttoned up? Is every fact in here cited to something real that we can verify? Like, what's our source of truth?

[Anne Morris]

And that's literally a question you would use as a prompt. Like, please make sure that every fact cited in this article has a verifiable source.

[Nick Lichtenberg]

Yeah. And it can be really simple, like include citations for all the facts and the quotes that you found here.

[Frances Fry]

Yep.

[Nick Lichtenberg]

And then it'll go back and it'll cite everything almost instantly. I was talking to Eric Brynjolfsson, who's a former mit, now Stanford AI thought leader, brilliant guy. And he said, I work with AI all the time and I'm asking it the wrong questions. It's what I'm learning. Like, it gives me back stuff I don't want. And then three seconds later it gives me exactly what I want. And it's because I didn't ask it the right question. And I was, I was confusing the AI with the way I was presenting information to it. I, um, and I talked, I talked to him about this and another CEO source of ours named Maria. And we all talked about how we find ourselves having conversations with AI while our spouse is confused in the other room. Like, what are you doing talking to a robot here? And it's because, I mean, from my perspective, if you layer in the right source of truth to begin with and then you add more context, you almost train the machine in terms of the right information, the path of information that you're looking for. And I think you can get to a good result and a good end place. But you have to be very intentional about the way everything is structured.

[Frances Fry]

I have no idea if what the practice I'm about to tell you if it helps or not, but I give a lot of positive reinforcement and constructive advice.

[Nick Lichtenberg]

Oh yeah.

[Frances Fry]

So I tell it, that was super helpful because this, this, and this. And then I say this one didn't work because of this. I'd like for you to try again, which is kind of how I talk to humans. And so, and I have found that it, it feels like it gets to what I want much more quickly since I've started giving it positive reinforcement and constructive advice in, in really detailed ways. Now I have no idea if I'm sure there's somebody out there testing these things to say whether or not it works. It's like in the beginning I was very polite to AI. It's just good practice for me to be polite. I don't know if that makes it better or not. It makes me feel better, puts me in a better mood when I'm interacting with it.

[Nick Lichtenberg]

No, I think you're doing a best practice because I think a mistake that I've made And I've heard other people make is that you really use these things to get what you want, but you're not ever telling them if they're doing a good job or not. And when you do tell them, this is a good example, this is a bad example, they will change their output to match the examples that you give them. And it doesn't have to be a ton of work for you. I think I've even taken screenshots of data and said, here are the results of what you gave me from the last couple months. And it instantly compresses all of that feedback and says, got it. Okay, now I have a better idea of what you want, and I'll try to give that to you, but then it'll still hallucinate something when you're least expecting it.

[Anne Morris]

Your editor in chief, Alison Chantel, told the Journal that your stories are 50% you. And we love a good arbitrary statistic. Why not 60%? Do you use your own thresholds like that as you kind of guide your way? And. And is that number different depending on the type of piece you're doing?

[Nick Lichtenberg]

Yeah, that number is different depending on the type of piece. So I've been able to do a ton of features where I get to sit down across the table from someone and I would use AI to help organize my thoughts. But I think, and it's a gut feeling and a judgment call, and maybe this will also be obsolete like 10 years from now, but I think that you have to take over and write that kind of thing out. And it's a question of news judgment on that one. I think, like a fun and stressful bit of this whole experience is that I do take pride in my writing, and writing means a lot to me. And I might be known forever as the guy who helped destroy writing at a pivotal moment in the news industry. And that's just beyond my control is what I would say. But it matters a lot to me that I write stuff that I'm proud of. And even the quick stuff where I am using AI to crunch a lot of notes and documents together. The 50% me that I think Alison was talking about, and I would defer. I'm not going to speak for her. 50% me is that the angle that you dive into the story, the personality that the. Even a dry news report can have, the sort of editorial spin that we put on the story, that's the 50% that comes from me. And I think that's why these stories have been read by millions of people at this point.

[Anne Morris]

There's a lot of hand wringing right now about how easy it is, allegedly. I think there's maybe some premise issues with what I'm about to say, how easy it is to allow AI to do our thinking for us. Like, you know, what will happen to the children, you know, sometimes literally right in like, this new world where they don't have to think ever. Is this something you worry about? What do you think are the downsides to this new way of working?

[Nick Lichtenberg]

I don't actually worry about that. I think this reminds me of growing up in the 80s and a lot of the talk about cable TV and explicit lyrics and in like hip hop music and heavy metal music and video games in the 90s. I think that there's always been a panic over, like, brain rot. I know brain rot is a more recent term, but I've been hearing about brain rot my whole life in different ways. And maybe, maybe we are like 40, 50 years into the brain rot era of modern culture, but I don't think so. I just think it's new and different. And I personally think from my experience that I've been working harder and using more of my brain than a lot of my career where this is a big challenge and you have to push yourself and push your AI tools to like, really do something that's good and that's not AI slop. And I think people settle for kind of the first output they get from AI and that's not what I do. I was talking to another professor about this and he said his students, he's concerned about them using AI to write their essays because they don't know yet what they want to say. But I feel like when I go into every article I'm putting together, knowing what I want to say and when I see what I get back, sometimes I know that's not it and that I have to keep pushing. So, yeah, it's a daily kind of pushing a rock up the hill for me. Like, I have to keep pushing my brain and my team and my AI co workers now to really say what we're trying to say, to really pull

[Frances Fry]

something off, you know, if I was only allowed to use one sentence to use it. Well, Nick, I think you just provided it, which is, does AI help you use more of your brain?

[Nick Lichtenberg]

Yeah. So I think we talked at the beginning about the fears that people have and the need for a pin cushion. It's not about the technology itself. And people are not trying, in my opinion, to really understand this technology. They're trying to grapple with that other fear that we're all going to stop using our brains or we're going to settle for using only part of our brains. And that's just like a totally different thing. I'm just trying to do a job,

[Anne Morris]

just trying to do my job. You've reflected that you were given a sandbox to play in and a lot of trust and freedom to kind of figure it out and that you think this moment probably still favors this sandbox approach. Do you stand by that observation?

[Nick Lichtenberg]

Yeah, I think if people have sort of sandboxes to play in and they can try to accomplish one specific goal, but they'll try lots of different things to get there, they will stumble upon something that works. And I hope that a lot of people are free to experiment and, you know, reimagine their job in a way that I think most people will be doing more interesting work after the end of this. I certainly have been. So that's what I'm hopeful about.

[Frances Fry]

I have been too.

[Anne Morris]

Say more Francis.

[Frances Fry]

Well, I spent a lot of time preparing to teach and thinking about how people learn and I might use like the really tight stylized example or a simulation or you know, a very famous for drawing a lot of things. Well, what I can do now with apps, it's just, it's like adding another dimension to what I can do and I'll draw something flat on the board and then I can give them an app where it comes to life for their example and I can feel, feel that more learning is occurring. Like if somebody said today, you know, data compression, whatever we're going to call it goes away, I'd be really sad because it would make me worse that it really has added a dimensionality to what is possible. And if I'm my favorite metric in the world is quality per unit time. Well, my quality per unit time in the classroom is much higher because because of AI than my quality per unit time in the classroom used to be.

[Capital One Sponsor]

This episode is brought to you by Capital One. Capital One's tech team isn't just talking about multi agentic AI. The they already deployed one. It's called Chat Concierge and it's simplifying car shopping using self reflection and layered reasoning with live API checks. It doesn't just help buyers find a car they love. It helps schedule a test drive, get pre approved for financing and estimate trade in value. Advanced, intuitive and deployed. That's how they stack. That's technology at Capital One.

[Progressive Sponsor]

You're listening to this podcast so I know you've got a curious mind. Here's a helpful fact you might not know yet. Drivers who switch and save with Progressive save over $900 on average. Pop over to progressive.com, answer some questions and you'll get a quick quote with discounts that are easy to come by. In fact, 99% of their auto customers earn at least one discount. Visit progressive.com and see if you can enjoy a little cash back. Progressive Casualty Insurance Company and affiliates national average 12 month savings by $946 by new customers surveyed who saved with Progressive between June 2024 and May 2025. Potential savings will vary. Support comes from Wise the smart way to manage the currencies you need around the globe Fed up with losing out to hidden fees when you send money abroad with your everyday bank? Choose the smart way Wise. You can count on the exchange rate you'd usually find on Google. No unwelcome surprises. Plus, ditch that where's my money feeling. Most transfers arrive in under 20 seconds. Join millions saving billions on hidden fees. Be smart, get wise. Download the Wise app today. T's and C's apply.

[Anne Morris]

Nick, you mentioned news judgment earlier. I'm curious what your advice is for people who are coming into the workforce right now figuring out how to use these tools without that news judgment. Like it seems like that's a big variable in how you're able to really like drive the Ferrari here.

[Nick Lichtenberg]

I wish I could remember where I read this when I was figuring out the declining industry I was getting into, but there was advice not to read one newspaper but to read four newspapers and then compare how they chose to assemble the facts. Like what did this one leave in and what did this one leave out and why did they do that? So you read four or five of them at the same time and you have your own opinion on what really happened in the world. And I think that's foreign to most people. I think most people read one story on one subject and they think they're done and they read the news and they're moving on. So I think media literacy is really investing in your own judgment what you think of the news based on what you're consuming.

[Anne Morris]

You wrote at one point that nobody knows anything is probably the most honest AI strategy right now. Do you have any predictions for where this goes?

[Nick Lichtenberg]

I think I'm pretty bearish on a lot of this. My own case notwithstanding. I don't think we'll see mass disruption of jobs. I don't even think we'll see mass disruption of industries. I think we'll see a very slow trickle of little bits of progress. And electricity is the metaphor that a lot of people are turning to. And it took 20 years for electricity to really be integrated. And I think AI will be like that, where some people think it's happening super fast. But we've seen with the corporate adoption struggles that it's just not that simple to rewire everything just like that, even though the tool is here and we can do it really quickly. So that's my prediction.

[Frances Fry]

I have an entirely different prediction because I get to watch the Nick Lichtenbergs of the world. And I think a lot of people are going to learn by extremes. And I think a lot of people are going to learn how to use AI by looking at your example. And I think it's going to go much faster.

[Anne Morris]

And predictably, I'm in between each of you. All right, Nick, I'd love to try to wrap a bow on this a little bit and give our listeners some who may still be skeptical of AI, like a pep talk for integrating these tools into their own work. So make the case for getting into that sandbox, regardless of industry and really experimenting with this stuff.

[Nick Lichtenberg]

Yeah, I would say, like, think of it as your assistant that's doing all the stuff that you don't want to do. Like Francis said, prepping for class is totally different now because you don't have to manually spend five hours putting something together. You collect the materials, you give them to your assistant, and they do it for you, and then you proofread them. And I think people are maybe clinging on to all the unpleasant stuff that they have to do because they're scared. They're scared of what's happening and they're scared that they're going to get dumber and that society is going to get dumber and bad things are going to happen. But the upside to me is you can spend more of your time doing higher level work. You create more time to have conversations like this one, because you have a robot assistant who is compiling spreadsheets for you and compiling tables of content for you and outlines for you and all these things that you just don't have to do anymore. So I would say look at it like that. It's a glass half empty, glass half full thing. Just like 50% nick. It's like glass half empty, 50% nick, or glass half full?

[Anne Morris]

I don't know. I love the. The fabulous assistant metaphor. Like, who doesn't want a fabulous, wildly competent assistant who knows all the historical facts.

[Nick Lichtenberg]

Yeah.

[Anne Morris]

Of the world? Like, that sounds very helpful to anyone.

[Frances Fry]

Make sure you give them feedback, they'll

[Nick Lichtenberg]

hallucinate that one thing and you'll be like, oh, I don't know about this assistant.

[Frances Fry]

They need a career path.

[Anne Morris]

90% is pretty good. Yeah, I love it. I love it. All right, final question we ask all our guests. What are you currently fixated on outside of work?

[Nick Lichtenberg]

So this episode may come out after it's done, but I have to say, the World cup has been such a fixation for me. I really love this sport. I've spent a good 20 years as a very passionate full time fan. I used to be one of those Americans who watched soccer once every four years when I was a kid. And then in my early 20s, I jumped into, started following the Arsenal Football Club and they just broke a streak. And I'm actually a Knicks fan too.

[Anne Morris]

So that's a good season for you.

[Nick Lichtenberg]

It's been a really surreal thing to have these two miracles happen four weeks apart. And now the World cup is just this traveling carnival and all the great stories about Europeans coming to America and saying, you know, we love the Buc ee's gas station in Texas and we love all the bars running out of beer in Boston, just like so many wonderful stories. And there was a lot of negativity coming into this World cup and it's been a surprisingly joyous experience. Not perfect, but joyous. So my wife is getting sick of me turning on the games every day, but it's hard for me not to fixate on it.

[Anne Morris]

Small price to play for joy at a global scale.

[Frances Fry]

It is a testimony to the power of sport that I don't think if anyone set out with the ambition to achieve what has been, they could possibly have done it. Like, this is, I think, the greatest testifying of the power of sport.

[Anne Morris]

Yeah, it's been pretty awesome in a literal sense.

[Nick Lichtenberg]

Yeah.

[Anne Morris]

We have two boys, 15 and 18, and they were just captivated by the spirit of the fans and the energy. And just like all of these humans, so comfortable in who they are and their passions. I mean, it just. There's so much feel good, shared humanity at a moment in time when I think we are all really hungry to be reminded that as a species we're fundamentally pretty good. Yeah, you feel good about Team Human. Speaking of our AI conversation, I do

[Nick Lichtenberg]

think AI will make everything feel more human because if everyone does use it for all the stuff they don't have to do anymore, like the 20th century will just get uploaded into this machine and we'll spend our time having more human experiences. I think that's the most hopeful thing about this technology. But I understand it's scary too.

[Frances Fry]

Sign me up.

[Anne Morris]

We're going to end on that note. That's a beautiful place to land. Nick, thank you so much for joining us. I love this conversation.

[Nick Lichtenberg]

Thank you. It was a real privilege to be asked on and thank you so much. That was fascinating.

[Anne Morris]

If you're enjoying this show, make sure to subscribe wherever you get your podcasts and and tell a friend to check us out. Fixable is a podcast brought to you by Ted and Pushkin Industries. It's hosted by me, Anne Morris and me, Frances Fry. This episode was produced by Trina Menino. Our team includes Constanza Gallardo, Banban Chang, Daniela Balaurasso and Roxanne Hylash.

[Frances Fry]

Our show is mixed by Louis at Storyyard.

[Capital One Sponsor]

This episode is brought to you by Capital One. Capital One's tech team isn't just talking about multi agentic AI. They they already deployed one. It's called Chat Concierge and it's simplifying car shopping using self reflection and layered reasoning with live API checks. It doesn't just help buyers find a car they love, it helps schedule a test drive, get pre approved for financing and estimate trade in value. Advanced, intuitive and deployed. That's how they stack. That's technology at Capital One.

[Nick Lichtenberg]

I love my phone but not my carrier.

[Capital One Sponsor]

What do I do?

[Anne Morris]

Well, there are 250 reasons to join T Mobile. Like what? You can keep your phone and your number and T Mobile helps pay it off up to $800 per line.

[Nick Lichtenberg]

So I wouldn't have to buy a new phone?

[Frances Fry]

Yep.

[Anne Morris]

Plus your plan price stays the same for 5 years on experience plans. Nice. Find your reason@t mobile.com

[Progressive Sponsor]

via virtual prepaid guard card typically takes 15 days after rebate submission price guarantee on talk text

[Nick Lichtenberg]

and 5G data exclusions like taxes and fees apply.

[Progressive Sponsor]

Lots of places can expose you to identity theft.

[Anne Morris]

Oh no.

[Progressive Sponsor]

That's why Lifelock monitors hundreds of millions of data points a second for threats to your identity. Which is way more than anyone can do on their own. If we find anything suspicious like new loans or changes to your financial accounts, we alert you right away. All through text, phone, email or the Lifelock app. Get the alerts that could make all the difference. Save up to 30% your first [e-mail] podcast terms apply.

[Frances Fry]

They say that money runs the world,

[Nick Lichtenberg]

but the story of money itself? That's where things get really interesting.

[Frances Fry]

Financial history has an uncanny way of repeating itself.

[Nick Lichtenberg]

How, you might ask. Well, we've got a brand new podcast that explains exactly how the Story of

[Anne Morris]

Money from the Financial Times. Hosted by me, Gillian, Ted and me, Robin Wigglesworth. Whether it's artificial intelligence bubbles or the

[Frances Fry]

next big market crash, if you don't

[Anne Morris]

understand the past, you just might lose

[Nick Lichtenberg]

money in the future.

[Frances Fry]

Follow the Story of Money wherever you get your podcasts.