In this compilation episode of The Parlor Room Presents: Hello AI, host and Harvard Business School Online Creative Director Chris Linnane gathers HBS faculty to share actionable advice for mid-career professionals navigating the AI landscape. Featuring professors Christina Wallace, Jake Cook, Iavor Bojinov, and Joe Fuller, the conversation explores how mid-career professionals can build AI fluency, apply their domain expertise, and create value in a rapidly evolving workplace.
From identifying entrepreneurial opportunities to experimenting with AI tools and leading organizational change, guests share what it takes to stay adaptable, relevant, and competitive in an AI-driven future.
Resources
Catch up on previous episodes of The Parlor Room, featuring faculty from this compilation episode:
You can also watch The Parlor Room on YouTube.
Transcript
Editor's Note: The following was prepared by a machine algorithm and may not perfectly reflect the interview's audio file.
Iav Bojinov: You have an opportunity right now to become that expert, to raise your hand and say, "I'm going to put in the time," because if you are the person who's driving the change, you become indispensable to the organization.
Voiceover: If you're in the middle of your career, how can you make AI work for you and not against you?
Jake Cook: You have to be willing to, regardless of whatever got you to this point, that old quote, "What got you here won't get you there," but use that to motivate yourself to get fluent with it.
Voiceover: Opportunities are everywhere if you can recognize them.
Christina Wallace: You don't have to raise money to get somewhere meaningful, even to the point where you could take it to a pilot customer and get some real revenue. It is so much easier to get started now. Your experience is your secret weapon.
Joe Fuller: You have a set of heuristics, a pretty good grasp of the way your world works, and that's really valuable.
Voiceover: Welcome to The Parlor Room Presents: Hello AI.
Chris Linnane: Hello, welcome to The Parlor Room Presents: Hello AI. Before we begin, I want to take a moment to thank you for listening to the show and to ask you a favor. If you could take a moment and like or follow the show wherever you get your podcasts, that would help us out a lot, and we'd really appreciate it.
Okay, so this is the last career advice episode of the season. We've released two early-career advice episodes and, two weeks ago, we released the first mid-career advice episode. This is the second career advice episode. On this episode, we speak with Christina Wallace, Jake Cook, Iav Bojinov, and Joe Fuller. Our first guest, Christina Wallace, gives us a kind of interesting perspective on mid-career advice, regarding AI, mixing in entrepreneurship.
So, if you're a mid-career professional, does AI now provide you with the opportunity to become an entrepreneur easier than in the past?
Christina Wallace: Yes, fundamentally. And as we were just talking about, what sets successful founders apart in this future world is proprietary data and insights. Mid-career professionals versus these 26-year-olds coming out of school have experience, have an understanding of a customer, have access to a network. They likely have proprietary data in a space that they have been working in, and using that can now go and build the good-enough with basically no resources now. You don't have to raise money to get somewhere meaningful with the first iteration of what you're building, even to the point where you could take it to a pilot customer and get some real revenue. It is so much easier to get started now.
And I liken it to if you look at in the 2000s, Pets.com, bless their hearts, they were selling pet food over the internet, and it was a total bust. And in 2017, Chewy comes out and sells pet food over the internet, and it's a huge successful public company. And you're like, same idea, what was different? Everything was different.
In the Pets.com era, you had to own your own servers, you had to hire your own operations and customer service folks, you had to store your own inventory and warehouse it, and then do the logistics of packing and shipping. You were processing credit cards in a moment when we didn't have great encryption standards, and customers were not comfortable sharing that information over the internet. So the behavior wasn't there, the trust wasn't there.
And then the cost structure of getting started was astronomical. 2017 comes around, and now you can build an e-commerce company on Shopify to get started. Are you going to scale it on Shopify? Probably not, but you can get started to see if there's a there there. You can outsource your inventory and your logistics to a third-party logistics provider. You can have different plugin tools that allow you to have access to customer service, even real-people customer service, but at a fractional level. You can get everything up and running in a day, and you get to take advantage of existing customer behavior and a safety-trust infrastructure that now exists that you didn't have to pay for.
And so I liken that shift to what we're seeing now where, well, even if you could build something on Shopify, you probably needed to find someone who understands Shopify because if you don't want just the plane templates and you want to tinker, now you're getting into the HTML and the CSS, and that's a little bit intimidating even if it's not that complex from a concept point of view. You can break things a lot easier than you can fix them when you don't know what you're doing. And you still need to hire four, five, six people once you are any sort of reasonable business because you can't do it all.
And I think now that has flattened even further to say you can't do it all eventually, maybe, but you can do a lot more. And as you think about it as a mid-career professional who's considering a shift, that means you could actually do a lot of this before you quit your job. Before you go all in, there's an ability to de-risk the opportunity much more than before to get a sense for, "Do I have an insight that stands out, or is my version of the world looking like everyone else's? Do I have traction? Do I have interest from this? Is this worth jumping for?"
Or, if you're trying something that has nothing to do with the world that you've been in, maybe you had the good-enough job for the last 20 years, and you're like, "What I really want to do is this." Great. Now use all of these tools to get up to speed on this industry you know nothing about, get started faster, and now use these tools to identify the experts that you're going to go ask and bring on as advisors and mentors.
Chris Linnane: Yeah. And what's great about this, what's so positive about this, is that's probably true for late career and retirees as well.
Christina Wallace: 100 percent.
Chris Linnane: Anyone who has an idea that they can now put into motion.
Christina Wallace: 100 percent. And as we think about the democratization of information, you also get the democratization of expertise, where, previously, I would go to my LinkedIn, and I would be looking among my network and my network's network for experts. And maybe I would go to the HBS alumni directory, and I have access to those things because of where I went to school and the rooms that I've been in up to this point. And so who I can reach out to for expertise is very different than the version of me that grew up in Michigan and stayed in Michigan, didn't go to business school, didn't spend time in Silicon Valley or in New York for over a decade and is trying to figure out, "Who do I go to to ask these questions that I don't want to ask the AI, that I want the expertise on?"
I can't promise you they're going to reply to the email, but now you know who they are. You probably are figuring out how to contact them because all of that is public, and now you have a shot of being able to connect to them and reach out and have that access. So I think it also, in some ways, devalues the exclusivity that these networks or this being in the room gave the elites, if you want to use that word. I think that opens up access to such a greater number of people.
Chris Linnane: One of my favorite discussions this season was with Jake Cook. In this clip, not only does Jake give us great career advice, mid-career advice in regards to AI, but he gives us an exercise to try. And I did try the exercise, and it worked out well. No spoilers. Here is Jake Cook.
When people are mid-career, and they ask you, "How do I stay relevant in this world of AI?" What kind of advice would you give them?
Jake Cook: If you're mid-career, you probably saw mobile and social at this point. And if, for example, I remember showing somebody a mid-career like Twitter, and they're like, "People post what they ate for breakfast. Who cares? I don't get it." And you can kind of dismiss this stuff. And disruption really does look like a toy on day one. It's just, "Give me a break, whatever."
So I think it was just very common in human behavior to become kind of set in your ways, and digital won't tolerate that. And so I think you have to be willing to, regardless of whatever got you to this point, that old quote, "What got you here won't get you there." I think if you were thinking that you're set and you can't be automated out, I think you should have a healthy dose of paranoia in a way to a point, don't be paralyzed by that, but use that to motivate yourself to get fluent with it.
And we did a conference for some Fortune 500 executives, and one of the exercises I did, and people were kind of like, raised eyebrow when I pitched this, like, "I think we should walk through a case, an example of using some data from customer service and show data hallucinating." And I'm going to walk them through live and prompt them. And people loved watching you prompt because I used the voice on ChatGPT. And the transcription's so good. And we're moving to this new interface that typing stuff to computers, we just talk to them like we would, which is great. And the LLM gets a lot more context because I'm too lazy to type.
Chris Linnane: Everything.
Jake Cook: Everything.
Chris Linnane: Yeah, you'll talk.
Jake Cook: And when you watch mid-career people that haven't maybe played with it too much, they treat it like Google. What's the best way to fix customer service issues? It's like, no, no, no, no, no, no, no, no. We have this new way where we can give the model a lot more. And the more context you give it, oftentimes the better it performs.
So mid-career, I would say, play with it in low-risk ways outside work to get you comfortable. And you do it with small experiments where you put skin in the game. And what I say is take the kind of Friday night challenge. You go to the store, and you pick whatever you're going to have here, get pizza delivered or whatever, it's Friday night, you don't want to cook, and pick a beverage or a dessert, and use AI to drive the purchase decision.
So you're learning a couple of things in that process. You're giving it context, I'm having pizza. I like this type of wine, or kombucha, or whatever that is, right?
Chris Linnane: Yeah.
Jake Cook: So now the model's like, "Oh, we're having cheese pizza with...oh, well, we don't necessarily want to have Mountain Dew."
Chris Linnane: Yeah, sure.
Jake Cook: "You didn't give me any clues; you wanted soda pop." Take a photo of the wine rack, for example, and say, "Based on these labels, which do you think is a good value?" Well, now the model has labels, it has prices. It can figure out based on that what the ratings are, what people have liked. So now we're using what they call multimodal, where we're taking imagery or video as an input for the model to help guide you, and you just iterate with it, and you play with it.
And then you have to kind of get to the end, and you're helping it kind of coach you, and then you have to take that beverage or dessert and put some money behind it. But it's a low-risk way, and you're learning prompting and where you can trust it and where the edges are. And then you're the ultimate customer at the end, like, "Did it give me a good result?" And I think these little micro experiments like that demystify this stuff.
Chris Linnane: Sure.
Jake Cook: It doesn't feel so daunting. And then you see how it may be, "God, I got this terrible bottle of white wine for a red sauce pizza, and it was awful." You can go through the rage and frustration of the robot lying to you.
Chris Linnane: But just on a Friday night instead...
Jake Cook: But on a Friday night.
Chris Linnane: ...of being in a meeting on Tuesday morning.
Jake Cook: Yeah, yeah. So you're out 15, 20 bucks on it or whatever. So I think those are really good ways to do that, and then start to kind of apply it in other areas as well.
Chris Linnane: Our next clip is with Iav Bojinov, who gives us a powerful perspective on how to apply your existing knowledge to AI to create true differentiation for the long term.
If you were advising a talented professional who feels uneasy about how fast AI is moving and it's changing their field, what would you tell them to do next?
Iav Bojinov: I would tell them to pause for a second and focus on AI, because it's not going to slow down. And what we've seen is that there is a growing gap between where companies are and where individuals are, which is not really evolving anywhere near as quickly as the rate of change and improvement of these AI models. And so you need to start today, you need to become proficient in these AI models, you need to start to understand how they're affecting your work, and you need to just start being a leader in this.
I think especially if you're sort of a mid-career professional, you've been in this area maybe 15 years, you know what you're doing. And so you have a huge competitive advantage because you have so much domain knowledge that all of the more junior employees who know how to use AI, because they all do now, they don't have that domain knowledge, they don't understand, they don't have that gut feeling, that intuition. And so you can actually start to perform so much better if you take the time to go back and figure out how to use AI to really help you in your work and to combine it with your specific domain knowledge. That's really the holy grail.
And every organization right now is looking for champions on AI. And this isn't going to happen overnight. I know it might seem like AI is moving so quickly. This is going to be the next five, 10, 15 years, maybe even longer. And so you have an opportunity right now to become that expert, to raise your hand and say, "I'm going to put in the time. I'm going to be the person who goes and speaks to all my other colleagues who are really scared of it, et cetera." Because if you're the person who's driving the change, you become indispensable to the organization. And so that's why I think it's important. If you haven't put in the time to learn about this, now is the time.
Let me do a little self-plug here. We did actually just launch a course with Professor Karim Lakhani on leading in the age of AI that's available on HBS Online. And so that course gives you all of the foundations that you would actually need to prepare yourself for this world where AI and agents are going to be ubiquitous in everything that you do. And so put in the time, learn this, and really cherish and be happy that you have all of this domain expertise that you have developed over the last 15 years, because that is those two things together are what's going to make you a rockstar.
Chris Linnane: That's perfect. When you think about it, you can't fake 15 years, but you can take a few months to get up to speed on AI.
Iav Bojinov: Exactly. And here's the thing, a lot of people really worry about, "I don't understand the latest thing. Oh, how do I know? What's the difference between ChatGPT 5.1 and 5.2?" That doesn't matter. What really matters, and that's what we've done in this course, is you need to understand some foundational things that are going to stay the same. So you need to learn those things, and then everything else is just small little changes.
So you need to, at a high level, understand how things like a transformer work, but you don't need to know what the latest iteration, the latest version of the algorithm. That's way too much detail. But you need to understand the implications for the work that you do, you need to understand the implications for your organization's business model, for the organization's operating model. Those are the things that you really need to understand. Those are staying the same at least for the next few years. And so yeah, pause, learn the foundations, and prepare yourself, step up, put your hand up, and be a leader. Don't try to hide in the sand because there won't be any sound left.
Chris Linnane: Now this last clip comes from Joe Fuller, and I saved it for last on purpose because this is practical advice, this is actionable advice. I would take his advice. So, here is Joe Fuller.
Joe Fuller: If you've been working in a function or an industry already for 15, 20 years of your life, you've got a lot of immensely valuable knowledge about what is generally described as a context. The context could be anything from what are the regulations related to personal health records because you're involved in health benefits in your employer, to how does the market for riding mowers work, to what would have been historically successful promotions for selling the seasonal product, barbecues, whatever? You have an understanding, what I'm going to describe as you have a set of heuristics which allow you to have a pretty good grasp of the way your world works, and that's really valuable.
So first of all, don't get discouraged. You actually got a very, very valuable currency in that contextual knowledge. The second thing is to accept that the world is going to change at a rate that is a little bit roller...it's going to be a little bit more like a rollercoaster. I know that's kind of a flip metaphor, but it's going to speed up, it's going to flow down, you're going to get dizzy. And I don't like rollercoasters.
Chris Linnane: Me neither.
Joe Fuller: But it doesn't mean I don't go on them with my grandchildren, but you're just going to have to accept that that's the circumstance. And the easiest way to settle your stomach is to embrace the types of changes that are happening generally under the rubric of generative AI and get comfortable with it. Don't wait for your employer to oblige you to do it.
Some of you will be aware that OpenAI's big launch product is called GPT. That stands for general-purpose technology. A general-purpose technology is one that's designed for near-universal application. The most common analogy that's made is to the deployment of electricity.
General-purpose technologies have attributes that make them universally applicable. To be a success, these AIs have had to be designed so that pretty much anyone can put them to use if they're willing to try.
My plea to you is, if you're not using it, start now. Start, by the way, getting on YouTube or one of the big large language model companies, websites like Anthropic or OpenAI. They offer free training on how to use it. There are courses that you can take from local universities. I think they're also very good, not terribly expensive, what we used to call MOOCs, massive online courses, where you can get a certificate, which, by the way, your employer might pay for with a benefit, or you forward it to your boss and you forward it to HR, and they'll say, "Oh, wow, George or Georgette really used some great initiative here." A lot of companies have programs where they do reimburse you, but only if you complete the course. So that's a good incentive.
But saying, "I'm scared of it," and/or, "I don't want to have to learn this." If you can type and you can talk, you know what you need to know to operate it. So I would be talking to peers about how they're using it. When I went to an industry event or a convention, I'd be doing that too.
You're also going to see that there is an unbelievable explosion right now in process or vertically-focused, what we call agentic AI. What's the difference between generative AI and agentic AI? Agentic AI, the word agentic comes from the word agent. And in decision-making processes, there's a phenomenon called the principal and the agent. The principal is the person who owns the asset or is in charge of the process, and they delegate to an agent making certain decisions and give them rules for making those decisions. So the board of directors of a public company are acting on behalf of the principals, who they're shareholders in that company, and they hire managers who are agents who try to create value for those shareholders.
The agentic AI is called agentic because it actually makes decisions. It's structured to take certain types of data. It's given decision rules to operate under that are instantiated in the logarithms that are inside the model that it's been trained on, and it makes decisions. The amount of venture capital going into focused agentic AI is, it's a tidal wave.
I was talking to a colleague who's super technically literate, and he said to me the other day, the other day, Monday, "If we wrote down what you were just saying using my language system," and then he named a large venture capitalist here in Boston, "And we could get a meeting there tomorrow. We'd leave with a check return of $5 million." I mean, there are 100 agentic AI companies focused on the legal industry alone. So there are going to be tools out there that are being designed specifically to help you find out what they are, try them out. If you're high enough up in the organization, maybe you call them up and say, "If you've got an alpha version, we'll be a test site for you."
But if you just demystify it, you're going to find something else as well, and you'll have a big...this is as big an incentive as I can offer you to take charge of your own future, which is a lot about your job you find boring and you can't stand and you don't want to do because you've done it 50 times, you've done it 100 times, and you don't learn anything, and it's routine, and it's tedious. So much of that can be made less onerous, less aggravating, less time-consuming through AI.
So if you dread writing that bimonthly report, if you dread having to take one source of data and reconcile it with another source of data, and then go to a meeting with people and argue about the interpretations of the data, a lot of that work can be advanced by AI. And it allows you to focus more time on your own skill development, on managing your people, on expanding your skill sets, on getting home at a more reasonable hour, and on not being worried about the next day's work. It really can be a big uplift to your job satisfaction, your quality of work life, once you get facile with it. And it's designed to make it easy for you to be facile with it.
Chris Linnane: So that concludes this episode of The Parlor Room Presents: Hello AI. Don't forget to like and/or follow wherever you get your podcasts. We really appreciate it. Thank you for listening.