Lenny Rachitsky: uncertainty

Transcript context

…I'm going to give you a little bit of my history with AI to get to that point. When AI was a completely different technology stack ... I have a master's in AI. So I started working in AI when ... It's evolved so much. And then of course at Google using it for auto complete and search. It's transformed so many times. But then with this last transformation of GenAI ... And that's what brought me to Glean of seeing this, meeting lots of AI companies and like, "Wow. This is really going to transform how we work." And it's just fascinating seeing these products. I was one of those people like, "Oh yeah, it's going to be so far away." Until I saw GPT3. And I think AI, we are underestimating how much it's going to change how we work. It's not going to be sudden from today to tomorrow because people haven't figured it out yet. They haven't figured out how exactly to leverage it. But the people who have are going to be so far ahead. They're going to be far ahead of everyone else because they're going to be working faster, they're going to be working better. And in five to 10 years, I think the lines between product managers and engineers and designers are going to blur because AI will enable product managers to build prototypes, to build designs. Designers to build a pro like Figma already has their Figma AI. You can press a button and you can get your initial prototype working. You've got all the co-pilots. So they're not quite there. You still, like with Copilot or with Cursor, you need to be an experienced engineer to know when it's getting it wrong, but they're going to keep getting better. I think people have to be careful about not getting left behind, but their jobs aren't going to go away. They're just going to change. I'm of the believer that we're just going to have a lot more software. But I talked to engineers and to PMs saying, "Yeah, I tried that. It doesn't really work." And then go back to how I worked before. And that's a dangerous spot to be in I think. For people that get anxiety hearing this where they're feeling they're going to be left behind and like, "Oh my God, I don't know enough time to do this or I don't know what I'm doing here. " Do you have any advice for what's something someone can do to not fall behind? Use the products. This is what good PMs should do period. Always be using new products. It's not a unique thing for AI. When mobile came out, PMs needed to be using mobile apps all the time to try them out, see what the UIs are, see what's working and what isn't. And the same goes even more for AI. Use ChatGPT. If your organization has Glean, use Glean. Use Claude. Try them all. Try them and see what they do. I was talking to a product manager I know who is more forward-thinking and loves playing with new products and he had this use case that blew me away. So he said ... And this was a couple of months ago, so before ... It was right when Gemini had expanded the context window. So his product had a Discord channel and he took the transcript from the Discord channel, which was huge. And he fed it into Gemini the entire channel and then used it to ask questions. Like what is the sentiment of my product? What is the most requested feature? What are the things people are unhappy with? This never would've occurred to me. It's like, that is so smart. And he's like, "It was like a goldmine." Do you know how long it would've taken him to read? And he just wouldn't have done it. So think about for the argument, oh, I'm too busy. Well, if you use these products it's going to be a leverage on your time. So you get a lot of articles sent to you, summarize them, use AI to summarize the articles. We use Gong at Glean to record all our sales calls. We have a Glean app that will read all the Gong transcripts, put them in a spreadsheet in a certain format of who the AE is, etc. And then summarize all the top requested features from all the Gong calls. And it took a while to get it right. At first the summarization, the prompt wasn't good enough and would give us features that our salespeople would recommend and didn't distinguish that this was actually the customer. So you have to tweak it. It's not going to work out of the box. But then we got it to the point where it worked and these things really save time and you have to use your creative juices as a PM to figure out how it can help you and have patience to iterate and keep trying because the models that we have today can do a lot already.…

Stored transcript either side of the excerpt. The highlighted words are the published quote; the surrounding text is unedited source, never generated.