YouSaid · the spoken record

Jeff Seibert

lines on the record
77
first
2023-11-22
most recent
2023-11-22
sittings or episodes
1
sources
podcast

Every line below is reproduced as it was said and linked to the record it came from. Nothing here is summarised or generated. Directory · Search · Corrections

  1. That is a really good question. So we built our ML team three years ago. We're training our own in-house models. You're right. Like you can't hallucinate in finance. We've done a ton of work to make sure our math is always accurate. It'll be really interesting to see every model has different strengths. Like if you already look at BARD and versus GPT and so on and so on, my sense is there'll be a very common popular open source sort of base LLM and then tools to allow folks to fine-tune it easily. What we've discovered is while it create it takes so much time data, energy, money to train an LLM, fine-tuning it can actually be done relatively straightforwardly with a surprisingly small amount of data as long as your data is very high quality and focused. And so imagine you basically have your sort of default Linux operating system, right? And then everyone customizes their flavor of it for their product, market use case, whatever it might be.

    2023-11-22 · The Twenty Minute VC · 20VC: Why OpenAI Will Become an Infrastructure Play, Why Apple Will Win in an AI World, Why Google is the Most Vulnerable Incumbent, Will LLMs Be Commoditised, Which Startups Are Thin vs Thick Wrappers on Top of LLMs with Jeff Seibert, Founder @ Digits · IDENTIFIED FROM THE TRANSCRIPT · source

  2. I don't know if it's different. So the closed may win in terms of having the most advanced model, but I think the Apple versus Android comparison is exactly accurate. So you're going to have something proprietary that might be best because it can be fully vertically integrated, that you control all the different variables. But I think there's going to be an open.

    2023-11-22 · The Twenty Minute VC · 20VC: Why OpenAI Will Become an Infrastructure Play, Why Apple Will Win in an AI World, Why Google is the Most Vulnerable Incumbent, Will LLMs Be Commoditised, Which Startups Are Thin vs Thick Wrappers on Top of LLMs with Jeff Seibert, Founder @ Digits · IDENTIFIED FROM THE TRANSCRIPT · source

  3. I certainly think we will. And this may not be a popular position. Obviously, OpenAI is charging ahead, sort of leading the way right now. I think the market force is at work mean there's just immense energy to have an open source equivalent. Meta appears to be highly motivated to open source its work. Many folks want to run these themselves and tune them themselves and so on. That is hard and expensive today, but I can't think of another thing in time in history where something hard and expensive in tech has lasted all that long. It's going to be commoditized.

    2023-11-22 · The Twenty Minute VC · 20VC: Why OpenAI Will Become an Infrastructure Play, Why Apple Will Win in an AI World, Why Google is the Most Vulnerable Incumbent, Will LLMs Be Commoditised, Which Startups Are Thin vs Thick Wrappers on Top of LLMs with Jeff Seibert, Founder @ Digits · IDENTIFIED FROM THE TRANSCRIPT · source

  4. Yet, this is such a special moment, and I'd say it vividly reminds me of the Rise of Mobile, which we wrote in Crash Lettix from 2010 to 2013, the Web 2.0 transition in 2005 to 2007. It's like these platform shifts are so rare, and I love them because that's where all the massive opportunities arise. And so the key, though, as a startup founder, is to be in a position to rapidly adapt to that new reality. And you need to be faster than your competition. So like for five years now, I've run digits on a weekly sprint. Every week we get to decide what our direction is that week. And I have to say that's been critical this year keeping up with all of the changes and evolution of the tech. No, I'm not particularly afraid of it. I love the energy of the tech world moving at a huge pace. And so the key thing, though, is this is technology. Like we are viewing it as a tool. It's not life or death for the company. It's like databases and so on. It's just evolving faster. Let's see how we use it and it gives us more capability.

    2023-11-22 · The Twenty Minute VC · 20VC: Why OpenAI Will Become an Infrastructure Play, Why Apple Will Win in an AI World, Why Google is the Most Vulnerable Incumbent, Will LLMs Be Commoditised, Which Startups Are Thin vs Thick Wrappers on Top of LLMs with Jeff Seibert, Founder @ Digits · IDENTIFIED FROM THE TRANSCRIPT · source

  5. The power of really deep intuition and conviction, looking at a market from a very theoretical level. So one of the most interesting aspects when he originally agreed to do our A round, I sort of asked him why afterwards, like why he committed so quickly. And he said, well, it had flashbacks to Uber because when Uber was going against the taxi industry, the NPS scores on taxis was so bad. Even if Uber was mediocre, it would still be way better. And he said accounting gave him the exact same vibes. The status quo is just so bad that if you can make accounting like somewhat enjoyable, it doesn't even need to be delightful. You've already won. And so his ability to distill these markets into these very high level, crisp, understandable talking points is super impressive.

    2023-11-22 · The Twenty Minute VC · 20VC: Why OpenAI Will Become an Infrastructure Play, Why Apple Will Win in an AI World, Why Google is the Most Vulnerable Incumbent, Will LLMs Be Commoditised, Which Startups Are Thin vs Thick Wrappers on Top of LLMs with Jeff Seibert, Founder @ Digits · IDENTIFIED FROM THE TRANSCRIPT · source

  6. 100%. And so, yeah, next to your team is obviously keep the investors informed and up to date, particularly your board. We're lucky to work with Peter Fenton at Benchmark. His clairvoyance on where the large opportunity is and just relentlessly pushing us to find that and adjust on a dime is super impressive. And I think hard to do. So definitely like keep your board tightly in the loop on this.

    2023-11-22 · The Twenty Minute VC · 20VC: Why OpenAI Will Become an Infrastructure Play, Why Apple Will Win in an AI World, Why Google is the Most Vulnerable Incumbent, Will LLMs Be Commoditised, Which Startups Are Thin vs Thick Wrappers on Top of LLMs with Jeff Seibert, Founder @ Digits · IDENTIFIED FROM THE TRANSCRIPT · source

  7. Think it's about this experiment thing, honestly. A bunch of angel investments I've made have tried to pivot, but I don't think they went all in on it. I think they saw it as a flyer that they'd try for a few months and they didn't see it as life or death. And by the time they realized it wasn't really working, it was life or death because they didn't have much cash left. And so you really need the conviction like day of to just sprint towards the new direction.

    2023-11-22 · The Twenty Minute VC · 20VC: Why OpenAI Will Become an Infrastructure Play, Why Apple Will Win in an AI World, Why Google is the Most Vulnerable Incumbent, Will LLMs Be Commoditised, Which Startups Are Thin vs Thick Wrappers on Top of LLMs with Jeff Seibert, Founder @ Digits · IDENTIFIED FROM THE TRANSCRIPT · source

  8. I agree with you. I think this is the hallmark of great founders you need to be able to convince your team and have the trust of your team to go all in on a new direction, disagreeing and committing in that scenario is like, okay, you might as well step back and like, let's just have a smaller team and really focus on this because it's not going to be productive.

    2023-11-22 · The Twenty Minute VC · 20VC: Why OpenAI Will Become an Infrastructure Play, Why Apple Will Win in an AI World, Why Google is the Most Vulnerable Incumbent, Will LLMs Be Commoditised, Which Startups Are Thin vs Thick Wrappers on Top of LLMs with Jeff Seibert, Founder @ Digits · IDENTIFIED FROM THE TRANSCRIPT · source

  9. Yeah, so the key is getting your team on board. If your team loses trust in you, you literally have no one to pivot, so it doesn't matter. They can really sense the uncertainty. My other advice, as I've said, is like, be very intentional and very decisive. And so both times we've pivoted digits, I gathered our core leadership team, laid out like, what are the challenges? What am I seeing? What are the options? We knew within 24 hours what the new path was and what the priorities were. We were all in on that new direction. It comes back to conviction. It is still a bet, but it's like, hey, here is the information we have on the field. We need to make a decision right now because what kills companies is uncertainty. And if you sort of muddle your priorities and have one team try this and another team try this, no one's heart is in it and both are going to be mediocre. I would rather see founders take like one to the moon bet on one new direction and it either works or doesn't.

    2023-11-22 · The Twenty Minute VC · 20VC: Why OpenAI Will Become an Infrastructure Play, Why Apple Will Win in an AI World, Why Google is the Most Vulnerable Incumbent, Will LLMs Be Commoditised, Which Startups Are Thin vs Thick Wrappers on Top of LLMs with Jeff Seibert, Founder @ Digits · IDENTIFIED FROM THE TRANSCRIPT · source

  10. It is not easy, but I think it's rare that you would have more than 12 months of cash because like usually you raise to have 18 months. And so by the time you figure out it's not going well and you need to pivot, it's 12, the big thing I would change from what you said is it's not an experiment. At no point where we're like, we're going to run two or three experiments. It is pivoting on a dime you are all in on the new direction and that is the only thing that matters to your success

    2023-11-22 · The Twenty Minute VC · 20VC: Why OpenAI Will Become an Infrastructure Play, Why Apple Will Win in an AI World, Why Google is the Most Vulnerable Incumbent, Will LLMs Be Commoditised, Which Startups Are Thin vs Thick Wrappers on Top of LLMs with Jeff Seibert, Founder @ Digits · IDENTIFIED FROM THE TRANSCRIPT · source

  11. That is the million dollar question. I'd say it's more of an art than a science. Like, you need to have a feel and an instinct as the founder of like, do you see a path to success? If sort of the windows closing on your

    2023-11-22 · The Twenty Minute VC · 20VC: Why OpenAI Will Become an Infrastructure Play, Why Apple Will Win in an AI World, Why Google is the Most Vulnerable Incumbent, Will LLMs Be Commoditised, Which Startups Are Thin vs Thick Wrappers on Top of LLMs with Jeff Seibert, Founder @ Digits · IDENTIFIED FROM THE TRANSCRIPT · source

  12. whoa hold on we're back like we can actually do what we set out to do literally overnight just like we're back focused on this and spent this whole year building it

    2023-11-22 · The Twenty Minute VC · 20VC: Why OpenAI Will Become an Infrastructure Play, Why Apple Will Win in an AI World, Why Google is the Most Vulnerable Incumbent, Will LLMs Be Commoditised, Which Startups Are Thin vs Thick Wrappers on Top of LLMs with Jeff Seibert, Founder @ Digits · IDENTIFIED FROM THE TRANSCRIPT · source

  13. Yeah, it's definitely been quite a journey. Obviously trying to make accounting real time is a lot easier said than done. When we started the company, we went heads down on R&D and really struggled with data quality for like three years. And that was because in 2018 when we started, the tech to really automate bookkeeping didn't fully exist. I think I was a little optimistic on how it could work. And so we have dozens of patents on it now, but it was like a brick wall. So in 2021, we made the decision to pivot from like the pure bookkeeping automation to collaboration tools. So better financial reporting, better client portals, better transaction review process. And that worked. We got a thousand accounting firms on the product, 5,000 downstream businesses. Like that was sort of off and running. But what bugged me is that wasn't really why we started the company. We had bigger ambitions. And so then last year, literally all of a sudden, GPT-3 comes out, chatGPT comes out, GPT-4, and we started experimenting and we're like,

    2023-11-22 · The Twenty Minute VC · 20VC: Why OpenAI Will Become an Infrastructure Play, Why Apple Will Win in an AI World, Why Google is the Most Vulnerable Incumbent, Will LLMs Be Commoditised, Which Startups Are Thin vs Thick Wrappers on Top of LLMs with Jeff Seibert, Founder @ Digits · IDENTIFIED FROM THE TRANSCRIPT · source

  14. Yeah, digits really came out of the crash lettuitics journey. And so, you know, we got very lucky with market timing. We scaled from 0 to 300 million phones in 12 months, got acquired by Twitter. Today, Crash Lytics is on five or six billion MAU, roughly every active smartphone on Earth. It's incredible. Through that journey, I was struck by this dichotomy. On the product side, you have real-time analytics, performance monitoring, live dashboards, right? Like I knew exactly what was going on with the product and who was using it. And then on the finance side, I literally had a black and white PDF of my P&L and balance sheet once a month, two to three weeks late that I didn't understand because I didn't have a background in finance. I was an engineer. And so it's like, what is happening? And so literally that is why I started digits. The simple premise, can we make accounting real time and intuitive for startup founders? And so what's crazy is it took us five years, but we finally just launched it like it's actually here five years later.

    2023-11-22 · The Twenty Minute VC · 20VC: Why OpenAI Will Become an Infrastructure Play, Why Apple Will Win in an AI World, Why Google is the Most Vulnerable Incumbent, Will LLMs Be Commoditised, Which Startups Are Thin vs Thick Wrappers on Top of LLMs with Jeff Seibert, Founder @ Digits · IDENTIFIED FROM THE TRANSCRIPT · source

  15. We do. It's also important to do it correctly. So I agree with your mentality. Crash Lytics, I never thought was like successful in any one moment, not when we were acquired, not when we hit a billion MAU, et cetera, et cetera. Like there's always the bigger goal. But if you have that mentality with the team, it's very demotivating. What are we trying to go to? Like, when are we going to get somewhere? And so it's really important to celebrate small wins. And so we use this Friday show and tell, we call it, basically, to show off what we did each week and champion who did what and celebrate all the small wins of the week so people feel really connected to the company and what's happening.

    2023-11-22 · The Twenty Minute VC · 20VC: Why OpenAI Will Become an Infrastructure Play, Why Apple Will Win in an AI World, Why Google is the Most Vulnerable Incumbent, Will LLMs Be Commoditised, Which Startups Are Thin vs Thick Wrappers on Top of LLMs with Jeff Seibert, Founder @ Digits · IDENTIFIED FROM THE TRANSCRIPT · source

  16. Yeah, it frackalizes. So, what happens is each team will run their retro on Friday and then surface sort of the highlights, like the biggest anchors or breezes to the full company-wide retro. Everyone has sort of two opportunities. You do a team-wide thing, and then you do your own small team, and that's where you get into more detail.

    2023-11-22 · The Twenty Minute VC · 20VC: Why OpenAI Will Become an Infrastructure Play, Why Apple Will Win in an AI World, Why Google is the Most Vulnerable Incumbent, Will LLMs Be Commoditised, Which Startups Are Thin vs Thick Wrappers on Top of LLMs with Jeff Seibert, Founder @ Digits · IDENTIFIED FROM THE TRANSCRIPT · source

  17. Yeah, it's certainly not easy because you can constantly fall into a trap of thinking you're getting feedback and you're not. It's really how you set the culture of the company. So one of the things we do at digits is we run the entire company on a weekly sprint as part of that every Friday we end every week with a full team retro. And we call it anchors and breezes. Anchors are what slowed you down, what didn't go well, what you need to help with, like feedback on the week, and then breezes are what went well. Shout outs to people who helped you, things you learned, etc, etc. And you create this culture of just constant iterative improvement, which then allows sort of feedback conversations and one-on-ones and so on to be widely recognized by the company is like, that's what we want. The whole mindset is just how do we get 1% better each week.

    2023-11-22 · The Twenty Minute VC · 20VC: Why OpenAI Will Become an Infrastructure Play, Why Apple Will Win in an AI World, Why Google is the Most Vulnerable Incumbent, Will LLMs Be Commoditised, Which Startups Are Thin vs Thick Wrappers on Top of LLMs with Jeff Seibert, Founder @ Digits · IDENTIFIED FROM THE TRANSCRIPT · source

  18. A hundred percent. And this is something I have felt strongly in for a long, long time. Google sort of pioneered this well in the early days and created this whole track for engineers to sort of keep climbing in comp and title and recognition and so on without taking on management roles. And I've tried to mimic that at all of my companies

    2023-11-22 · The Twenty Minute VC · 20VC: Why OpenAI Will Become an Infrastructure Play, Why Apple Will Win in an AI World, Why Google is the Most Vulnerable Incumbent, Will LLMs Be Commoditised, Which Startups Are Thin vs Thick Wrappers on Top of LLMs with Jeff Seibert, Founder @ Digits · IDENTIFIED FROM THE TRANSCRIPT · source

  19. That is a great question. So, actually, every time I've promoted an IC to a manager in a startup, this was not a Twitter, we've done it as sort of a trial period. And so it's like, hey, so-and-so, we have this opening for this new role. The team's growing. We need some structure. We are going to have so-and-so take on the role as a trial over the next two months. And let's see how they do. And honestly, there's been both outcomes. In some circumstances, it's been great and everyone's rallied around them. It's like, great, okay, now they're the manager for that team. And I've had circumstances where they haven't. And it was sort of widely recognized that they weren't excelling in that role. And we decided to move them back to an IC. And that was okay as well. I think there's a really bad perception that being a manager is better than being an IC. I think it is different. And you can be an exceptional IC and you should be compropriately for that. Or if your career passion is to mentor and guide folks, then you move into management.

    2023-11-22 · The Twenty Minute VC · 20VC: Why OpenAI Will Become an Infrastructure Play, Why Apple Will Win in an AI World, Why Google is the Most Vulnerable Incumbent, Will LLMs Be Commoditised, Which Startups Are Thin vs Thick Wrappers on Top of LLMs with Jeff Seibert, Founder @ Digits · IDENTIFIED FROM THE TRANSCRIPT · source

  20. Peter principal they get promoted into management because they were good at a former job their passion, right? Their experience isn't managing. The feedback cycle is slow. They're the boss so they don't get the raw feedback. I think there's a ton of challenges and it's even worse for CEOs because who in your company is going to give you really crisp blunt feedback on what you did right and wrong? And your investors aren't involved enough in the day-to-day to really know. So they can give strategy advice, but they don't know how you're behaving in meetings.

    2023-11-22 · The Twenty Minute VC · 20VC: Why OpenAI Will Become an Infrastructure Play, Why Apple Will Win in an AI World, Why Google is the Most Vulnerable Incumbent, Will LLMs Be Commoditised, Which Startups Are Thin vs Thick Wrappers on Top of LLMs with Jeff Seibert, Founder @ Digits · IDENTIFIED FROM THE TRANSCRIPT · source

  21. This sounds silly but honestly pure execution. People know the vast majority of managers are terrible managers, right? The vast majority of founders are simply bad at running companies. And I'm sorry, but it's true. And so what I mean by that is like most founders aren't intentional.

    2023-11-22 · The Twenty Minute VC · 20VC: Why OpenAI Will Become an Infrastructure Play, Why Apple Will Win in an AI World, Why Google is the Most Vulnerable Incumbent, Will LLMs Be Commoditised, Which Startups Are Thin vs Thick Wrappers on Top of LLMs with Jeff Seibert, Founder @ Digits · IDENTIFIED FROM THE TRANSCRIPT · source

  22. I think it is critical and it is too easy for companies to fall into like, oh, we don't know, we need more data. We need to run a survey for that, particularly at Twitter. We need to get statsig data, which always took two to three weeks. And then you would have to analyze that and then see what happened. It's honestly a disaster. And so agree or disagree with some of Elon's decisions, but he is moving quickly in a direction that is way better than sort of standing still.

    2023-11-22 · The Twenty Minute VC · 20VC: Why OpenAI Will Become an Infrastructure Play, Why Apple Will Win in an AI World, Why Google is the Most Vulnerable Incumbent, Will LLMs Be Commoditised, Which Startups Are Thin vs Thick Wrappers on Top of LLMs with Jeff Seibert, Founder @ Digits · IDENTIFIED FROM THE TRANSCRIPT · source

  23. I'm definitely on the side of build it for yourself. Like you deeply understand the problem, that gives you superpowers in terms of solving it the right way. You definitely need to understand, are you repeated by many other people around the world? Do they share your problem? Do they share how you think it should be solved? But I think it's so much easier to build something that you personally feel than have to sort of try to interpret other people's thoughts and beliefs about it. So no, I would not be worried about getting too emotionally involved. I think that's a superpower.

    2023-11-22 · The Twenty Minute VC · 20VC: Why OpenAI Will Become an Infrastructure Play, Why Apple Will Win in an AI World, Why Google is the Most Vulnerable Incumbent, Will LLMs Be Commoditised, Which Startups Are Thin vs Thick Wrappers on Top of LLMs with Jeff Seibert, Founder @ Digits · IDENTIFIED FROM THE TRANSCRIPT · source

  24. Yeah, the biggest lesson I learned was empathy, honestly. So when you're in consumer software, you can't possibly begin to understand how many different people, personas, use cases, mindsets, the human experience all comes to bear on your product. And what I saw was actually a trap. So the product managers who were super data-driven started designing and building features for the average user, because that's what the data told them. And they actually believed that there was something such as an average Twitter user. It's such a huge mistake, right? Like you're conflating all of these different populations. You have sports fans who want a live chronological timeline during the game. You have celebrities who want to maximize their reach. You have Japanese users who by and large want to remain anonymous. None of them is average. And so what I really learned is you have to deeply understand each population and design and build a feature for them. Don't like let the data lie to you.

    2023-11-22 · The Twenty Minute VC · 20VC: Why OpenAI Will Become an Infrastructure Play, Why Apple Will Win in an AI World, Why Google is the Most Vulnerable Incumbent, Will LLMs Be Commoditised, Which Startups Are Thin vs Thick Wrappers on Top of LLMs with Jeff Seibert, Founder @ Digits · IDENTIFIED FROM THE TRANSCRIPT · source

  25. Man, so I loved building things since I was little, and I was completely obsessed with Legos. And so my dream was honestly literally to be a Lego master builder until my mom did some research and she found out that actually like it's not that great of a career. They're paid something like minimum wage. And that was in middle school. And so I forget if it was that Christmas or the next year, but she gave me a programming book for Christmas. And that was the end of the story. I was like, okay, computers are next.

    2023-11-22 · The Twenty Minute VC · 20VC: Why OpenAI Will Become an Infrastructure Play, Why Apple Will Win in an AI World, Why Google is the Most Vulnerable Incumbent, Will LLMs Be Commoditised, Which Startups Are Thin vs Thick Wrappers on Top of LLMs with Jeff Seibert, Founder @ Digits · IDENTIFIED FROM THE TRANSCRIPT · source

  26. Harry, no, it is so great to be here. Thanks for having me on. You know, the whole world only makes sense going backwards because, like, who would have guessed from that meeting you'd become one of my largest investors like five years later? Just incredible what you've done.

    2023-11-22 · The Twenty Minute VC · 20VC: Why OpenAI Will Become an Infrastructure Play, Why Apple Will Win in an AI World, Why Google is the Most Vulnerable Incumbent, Will LLMs Be Commoditised, Which Startups Are Thin vs Thick Wrappers on Top of LLMs with Jeff Seibert, Founder @ Digits · IDENTIFIED FROM THE TRANSCRIPT · source

  27. I view OpenAI probably evolving more into an infrastructure company like AWS. The road ahead for OpenAI is not easy. Google, they need to go all in on it. I don't think they have a choice. Very few people I think are paying attention to is Apple because again, they control the silicon. Imagine they're able to pioneer small models that run on device and then they do custom silicon to make them run. The performance could be outlandish compared to any other platform.

    2023-11-22 · The Twenty Minute VC · 20VC: Why OpenAI Will Become an Infrastructure Play, Why Apple Will Win in an AI World, Why Google is the Most Vulnerable Incumbent, Will LLMs Be Commoditised, Which Startups Are Thin vs Thick Wrappers on Top of LLMs with Jeff Seibert, Founder @ Digits · IDENTIFIED FROM THE TRANSCRIPT · source