Source:Silicon Valley Girl
Compiled by:Felix, PANews
Summary: Any work that can be called "craftsmanship" is not easily replaced, as AI is not good at handling the nuances within.
Bill Gurley has witnessed the rise and fall of countless careers over 25 years. The companies he has invested in are worth over $50 billion. Those who lost everything were not the risk-takers, but those who chose to play it safe.
On the Silicon Valley Girl podcast, Bill Gurley primarily discussed how AI is changing work, which professions will disappear first, and what actions to take if jobs are already starting to shrink. PANews has compiled highlights from the conversation.
Host: I originally had a question prepared, but news just broke that Block is laying off nearly half its employees to boldly embrace AI. Should people be scared?
Bill: Yes, but with some qualification. AI tools have been around for two or three years now; they won't be taken back, and they won't disappear. For anyone in any field, the most direct advice to protect yourself from the impact of AI is: Strive to become the 'AI-empowered' version of yourself. You must understand the boundaries of AI capabilities in your field and use it as much as possible. The more you use it, the more types of prompts you'll think of, and you'll begin to understand its capabilities. If you don't proactively leverage it, you will be left far behind by those who do.
The novelty is that it's impacting white-collar jobs, which were previously considered "safe." If you are actually replaced, you can ask yourself: Is this really the work you love and want to do? If not, this might be your opportunity to go find what you truly love.
Host: We used to think IQ would help us sail through life. But you're now saying "Playing it safe / seeking stability is now the riskiest move." Can you explain?
Bill: I think the sad reality is that many career advisors, counselors, and parents often push people towards jobs they perceive as "safe." But if you don't love the work, according to a 2023 Gallup poll, over 50% of people are not actively engaged at work, with only about 23% truly "engaged." If you are in a position you don't care about and have no motivation to improve yourself, you are a sitting duck in the face of these changes. Parents' intentions are for their children's financial stability, but you only live once. If you can find something you are intensely curious about, you can create a huge gap between yourself and others, and the financial rewards will follow.
Host: Can you list three characteristics of people who don't "play it safe"? What do they do every day?
Bill: First, allow yourself to chase careers you once thought were impossible. For example, Shake Shack founder Danny Meyer was going to law school to become a lawyer, but encouraged by his uncle, he confronted his childhood hobby of finding good restaurants on weekends and writing diaries, and eventually went into the restaurant business. Second, hone your craft, keep learning. If you love something, you will continuously learn out of curiosity; absorbing new information will energize you, while being forced to take classes in a field you dislike will only bore you and drain your energy. Finally, be at the knowledge frontier. AI and large language models record the best practices that have already been written down, but the frontier knowledge being explored today is not yet in the models. If you are the one mastering frontier knowledge, you are ahead of the AI model.
Host: When evaluating founders, what kind of behavior do you consider as not "playing it safe"?
Bill: Actually, I think the entire venture capital world has recognized that you absolutely cannot have a "play it safe" mentality when choosing founders. I think the best and brightest are very independent thinkers. They naturally think outside the box. If some people find them a bit "crazy" because they are so dogmatically firm, that's actually right too. And I don't think I'm the only one who would say this. If you study essays by venture capitalists, you'll find almost a trend of looking for these personality traits. This goes back to that Jobs ad about the "crazy ones."
Host: It feels like everyone should adopt this "crazy" mindset to succeed in this world now, because you are becoming the entrepreneur of your own life, right?
Bill: Yes. I think that's a good way to put it. Especially in certain fields where this new technology is causing so much disruption, if you don't take control of your career path, you are likely at risk. I often say, if you are a person with high agency, confident in driving your own career development, these tools are actually a "jetpack." If you want to learn proactively, there has never been a time in history when you could learn this fast, except with AI. And podcasts, like the shows you do. YouTube is full of interviews; you can learn at an unprecedented speed. This is also part of the reason why I have little reaction to "AI anxiety," because anxiety might freeze you, and what you should do now is run as fast as you can.
Host: Sometimes when I'm scrolling on my phone, I hear people, including you, say: We have a window of opportunity that is closing, so we should work very hard now to seize it. Can you elaborate? How much time do we have? How hard should we work?
Bill: I don't think the window is necessarily closing, it's just that this tool came so fast, with so many functions, everyone must figure out as quickly as possible what it can do in their industry. I don't like those dystopian doomsday narratives; Dalio is arguably the worst doomer, and he always has the microphone. I don't think that's healthy. This anxiety is about 5 times higher in the US than in China. While this anxiety can sometimes stimulate people to act, it's more likely to paralyze them. Instead of being anxious, what you need to do now is run as fast as you can.
Host: You also mentioned that it's very important to "forget/abandon what made you successful." How do you do that?
Bill: There's a saying: "Strong opinions, weakly held." It's hard to act without strong opinions, but never treat any opinion as sacrosanct or immutable. If you develop the habit of continuous learning and realize the risk of holding on to past views, you can best judge when to let go of past experiences.
Host: Faced with so many curious things, how do you find a direction that can become a lifelong career? What if you choose wrong?
Bill: You can try to turn a "hobby" into a career. For example, I worked as an engineer for two and a half years, then went home at night and read Peter Lynch's "Beating the Street," started researching and trading stocks—that's what fascinated me, and later I went to work on Wall Street. Don't feel trapped because you don't know the answer; many people don't find the perfect fit until they are 30 or 40. You need to create more "shots on goal" for yourself and reflect once a year: "Is this what I want to do for the rest of my life? Will I still want to do this in 30 years?" Once the answer is "no," you should look for other directions. Bezos's "regret minimization framework" is very similar: imagine what your 80-year-old self would advise about your current career choice. We did a survey; if they could go back and start over, 6 out of 10 people said they would make a different career choice. Many choose wrong because they listened to a professor's praise or their parents' advice for economic stability and became doctors or lawyers, without asking themselves what they truly love.
Host: Can you list a few jobs that can resist AI impact, and jobs that are about to be eliminated?
Bill: The first to be threatened are jobs dealing with language, such as translators, paralegals, etc., because large language models are very good at reorganizing text. Also, if your job is just mechanically writing code day in and day out, demand will decrease because code is essentially a more constrained language than natural language. If you are the type who knows how to judge structurally why one algorithm is better than another, or knows how to streamline code to make it more efficient, I think these skills will still be valuable. But I must return to the previous topic: the best way to be a good software engineer in the AI world is to be the one who knows how all the new AI tools work. Be the person in your organization who actively embraces them, not rejects them.
Jobs that are more resistant to AI are any that can be called "craftsmanship," because they deeply understand the "nuances" of their field, and AI is not good at handling nuances. Additionally, interpersonal relationships will become more important. Do you have a strong peer network? A strong mentor network? A strong relationship network. I think this is very helpful.
Host: What should ordinary people do now to take their AI tool setup to the next level?
Bill: The most important thing is to try various mainstream models, like ChatGPT, Claude, Gemini, etc., and let AI take on more agentic work. For example, before a podcast interview, I ask AI to predict what questions the host will ask me; or when preparing a TED talk, I use AI to simultaneously conduct research, prototype, and brainstorm ideas. Many people don't realize how much they can rely on AI. When I spoke at NYU, I found that half the questions students asked could have been asked directly to ChatGPT. Now for any information query, it's much easier to ask AI directly than to look it up on Wikipedia.
Host: How to find mentors and build a valuable peer network?
Bill: Regarding mentors, most people aim too high, going directly to people who are unlikely to pay attention to you. I suggest two strategies. First, create a mentor list. Use free online resources (podcasts, interviews, AI queries) to research people you admire, building a digital dossier on them like a fan following a star. This not only builds your confidence but also tests whether you are truly passionate about this field.
Furthermore, find people a step or two down the ladder to be your actual mentors. If they've never been asked to be a mentor before, they will be flattered. Don't start with a serious question like "Can you be my mentor?" but make a specific, small request, like "I'm researching these two paths, you must have considered them, any advice?" Or you could simply use your personal corpus of data to create an AI project and use it as a virtual mentor for conversation.
Regarding peer networks, I suggest that throughout your career, find 4 to 6 like-minded people at a similar stage (preferably outside your organization) and create a group on WhatsApp or Slack, etc. You can share ideas, expand your learning, and broaden your network. When you have a bad day at work or are confused about whether the problem is the industry itself or your current company, this group can help you see clearly.
Host: In this era, how should parents educate their children? Are cram schools still useful?
Bill: There is a serious "résumé race" in North America now; parents even start worrying about their children's college applications from sixth grade, filling their schedules with chess, lacrosse, even volunteering at animal shelters. While it's good to teach children perseverance, I worry that children will be completely exhausted by the end of college. If they never have time to stop and breathe to explore and discover what they really want to do, that's terrible. Now universities even require declaring a major early, robbing them of the opportunity to explore. I advise parents to create as many exploration opportunities as possible for children, expose them to different things, and try to discover their true, pure interests.
Host: For people doing jobs they are afraid to quit and change, what is one thing they can do this week?
Bill: You can conduct scenario planning for new paths in the digital world. Set a scenario: "I plan to leave here in six months," then use AI tools to help you simulate it.
For example, you can ask AI: "I want to switch to this field in six months, what should my first week look like?" It will give you a plan. You don't have to follow exactly what AI says; you can modify it, but through this scenario planning process, it can help you break out of the paralysis. You can do this for three different directions, enriching it with new data each week. Once the details are filled in, it's no longer an abstract concept, and you'll start to see which path you lean towards.
Related reading:Transcript of Jensen Huang's Latest Podcast: Nvidia's Future, "AI Doomsday" Theory, Corporate Moats...
