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

In this episode of 20 VC, host Harry Stebbings converses with Jeff Seibert, a seasoned entrepreneur and founder of Digits, a company transforming accounting with AI and backed by investors like Peter Fenton and Benchmark. They delve into the potential of AI in revolutionizing industries, with Seibert highlighting OpenAI's possible transition into an infrastructure company akin to AWS. Seibert reflects on his journey from a Lego-obsessed child to a tech entrepreneur, emphasizing the importance of empathy in product management and the pitfalls of designing for the "average user." They discuss the necessity for startups to pivot decisively based on founder conviction and market understanding, rather than data alone. Seibert also shares insights from his angel investing experiences, stressing the unpredictability of startup success and the value of staying customer-focused over being preoccupied with competition. The conversation touches on the rapid adoption of AI, the commoditization of large language models (LLMs), and the implications of climate change, offering a candid look at the challenges and opportunities in the tech landscape.

• Jeff Seibert views OpenAI as potentially evolving into an infrastructure company similar to AWS.
• He acknowledges the challenging road ahead for OpenAI, especially with competition from Google.
• Jeff emphasizes the importance of Apple in the AI space due to their control over silicon and the potential for pioneering small models that run on device with custom silicon.

"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."

This quote underlines Jeff's perspective on OpenAI's future and the competitive landscape in AI, highlighting Google's imperative to invest heavily in AI.

"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."

Jeff points out Apple's unique position in the AI race due to their control over hardware, which could lead to significant advancements in on-device AI performance.

• Harry Stebbings shares his personal connection with Jeff, beginning with their meeting nine years ago at Twitter.
• Harry's friendship with Jeff evolved into an angel investment and eventually, 20 VC became one of the largest investors in Jeff's company, Digits.
• Harry introduces Jeff Seibert, highlighting his past successes, including co-founding Crashlytics and his role at Twitter before founding Digits.

"Now this show is an immensely special one for me. I first met this guest nine years ago when he was head of consumer product at Twitter."

Harry expresses the personal significance of the episode, referencing his long-standing relationship with Jeff.

"That friendship with today's guest then turned into an angel investment from me into his company, and today, 20 VC is one of today's guests and his company's largest investors."

This quote details the progression of Harry's relationship with Jeff from friendship to financial investment and support.

• Jeff Seibert was passionate about building things since childhood and was obsessed with Legos.
• His mother introduced him to programming after learning about the limited career prospects of a Lego master builder.
• Jeff's early interest in computers set the stage for his future in technology and entrepreneurship.

"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, it's not that great of a career."

Jeff describes his initial dream of becoming a Lego master builder and how his mother's research led to a pivot towards programming.

• Jeff Seibert learned the importance of empathy during his time at Twitter.
• He observed the pitfalls of designing for an "average user" and the necessity of understanding different user populations.
• Jeff emphasizes the importance of being intentional in operations and not relying solely on data.

"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."

Jeff reflects on the crucial lesson of empathy he learned at Twitter, which shaped his understanding of diverse user needs.

"What I really learned is you have to deeply understand each population and design and build a feature for them. Don't let the data lie to you."

He warns against the dangers of designing features based on data that represents an "average user," which doesn't truly exist.

• Jeff Seibert advocates for building products based on personal experience, believing it provides a deep understanding of the problem and a powerful motivation.
• He acknowledges the need to verify that others share the same problem and solution perspective.

"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."

Jeff expresses his stance on the debate between building products based on personal experience versus detachment, favoring the former for its advantages in problem-solving.

• Jeff Seibert considers speed to be critical in product development, highlighting the risks of over-reliance on data and surveys.
• He appreciates the quick decision-making and direction seen at Twitter under Elon Musk's leadership.

"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."

Jeff criticizes the hesitation and indecisiveness that can slow down product development, emphasizing the importance of speed.

• Jeff Seibert identifies execution as the most misunderstood element of entrepreneurship.
• He points out that many founders are not intentional in their operations and decision-making.
• Jeff emphasizes the need for founders to have conviction and learn from their intentional decisions.

"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."

Jeff discusses the common shortcomings of managers and founders, stressing the importance of effective execution.

• Jeff Seibert discusses the Peter Principle and the challenges of providing feedback to managers and CEOs.
• He describes his approach to promoting individual contributors (ICs) to management roles as a trial period.
• Jeff advocates for recognizing the value of exceptional ICs and compensating them appropriately without forcing them into management.

"Peter? Principle. They get promoted into management because they were good at a former job. They're passion, right? Their experience isn't managing."

Jeff explains how the Peter Principle leads to poor management by promoting individuals based on prior success rather than management skills.

• Jeff Seibert talks about creating a culture of accountability and iterative improvement within a company.
• He describes the weekly sprint system at Digits, which includes a full team retrospective to discuss both successes and areas for improvement.
• Jeff believes in celebrating small wins to motivate the team while maintaining a focus on larger goals.

"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."

Jeff shares how the culture at Digits is shaped by a weekly sprint cycle that fosters continuous feedback and improvement.

• Jeff Seibert's experience at Crashlytics, which scaled rapidly and was acquired by Twitter, led to the founding of Digits.
• He was inspired to start Digits due to the contrast between real-time product analytics and delayed, confusing financial reports.
• Digits aims to make accounting real-time and intuitive for startup founders, a goal that took five years to achieve.

"And so literally, that is why I started digits. The simple premise, can we make accounting real time and intuitive for startup founders?"

Jeff explains the motivation behind founding Digits, which was born out of his frustrations with traditional accounting practices.

• Jeff Seibert discusses the challenges faced in the initial R&D phase of Digits, particularly with data quality.
• The initial focus on bookkeeping automation proved difficult, leading to a pivot towards collaboration tools and financial reporting.
• Jeff emphasizes the importance of being adaptable and responsive to challenges in the entrepreneurial journey.

"When we started the company, we went heads down on r and D and really struggled with data quality for like three years."

This quote highlights the initial struggles with data quality that Digits faced, which eventually led to a strategic pivot in the company's focus.## Initial Pivot and Company Direction

• Jeff Seibert's company initially gained traction with 1000 accounting firms and 5000 downstream businesses using their transaction review process.
• The company had larger ambitions beyond their initial success.
• The advent of GPT-3, chat GPT, and GPT-4 reignited the company's original vision, allowing them to pursue their initial goals.

"And that worked. We got 1000 accounting firms on the product, 5000 downstream businesses. 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." "GPT-3 comes out, chat, GBT comes out, GPT four. And we started experimenting and we're like, whoa, hold on, we're back."

The quotes highlight the company's shift from their initial product offering to refocusing on their original, more ambitious goals due to new technological advancements.

• Pivoting is more art than science, requiring founder instinct.
• A path to success must be visible; otherwise, a pivot may be necessary.
• Many successful companies resulted from hard pivots.
• Founders should have at least a year of cash left to see a pivot through.

"I'd say it's more of an art than a science." "That's, to me, when you have to go pivot." "You need at least a year of cash left, because if you don't have a year of cash, you're not going to have time to see this pivot through."

These quotes encapsulate the delicate balance of knowing when to pivot and the practical considerations such as financial runway needed to implement a pivot successfully.

• Conviction and decisiveness are critical during a pivot.
• The team's trust is paramount; without it, a pivot is impossible.
• Pivots should not be treated as experiments but as all-in commitments.

"The key is getting your team on board." "be very intentional and very decisive." "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."

Jeff Seibert highlights the importance of having a committed team and the necessity of founder conviction when making a pivot.

• Keeping investors and the board informed is crucial during a pivot.
• Investors often back founders they believe in, even through uncertainty.

"And so, yeah, next to your team is obviously keep the investors informed and up to date, particularly your board."

Jeff Seibert emphasizes the importance of maintaining open communication with investors and board members during significant company shifts.

• Peter Fenton's deep intuition and conviction are key lessons.
• The ability to distill markets into high-level, understandable points is valuable.

"The power of really deep intuition and conviction." "his ability to distill these markets into these very high level, crisp, understandable talking points is super impressive."

Jeff Seibert shares his admiration for Peter Fenton's approach to understanding markets and making investment decisions.

• Adapting to platform shifts rapidly is crucial for startups.
• Viewing technology as a tool, not as a dependency, is essential.
• Startups must focus on solving customer problems first and foremost.

"This is such a special moment." "It's like these platform shifts are so rare, and I love them because that's where all the massive opportunities arise."

Jeff Seibert compares the rise of AI technologies to previous platform shifts and stresses the importance of agility in leveraging these technologies.

• Open source equivalents to proprietary LLMs are likely to emerge.
• Commoditization of technology is a common trend.

"It's going to be commoditized."

Jeff Seibert predicts the inevitable commoditization of LLMs, drawing parallels to past trends in technology.

• Closed systems may have the most advanced models, but open systems will be close seconds.
• The market may see a dynamic similar to Apple vs. Android in AI.

"But I think there's going to be an open source equivalent or more open equivalent. That's a very close second."

Jeff Seibert discusses the potential for open AI systems to coexist with closed, proprietary systems, offering alternatives to users.

• Specialization of LLMs for different use cases is anticipated.
• Companies may leverage multiple LLMs, each with different strengths.

"Every model has different strengths, like if you already look at bard versus GPT and so on and so on."

Jeff Seibert talks about the diversity of strengths across different LLMs, suggesting a future where companies might use multiple models for different purposes.

• Quality of data is crucial for fine-tuning models.
• The base LLM layer's performance is correlated with data size, but fine-tuning may not require large data sets.

"What I'm talking about is sort of the next tier of how do you fine tune the models? And that's where actually, I think the quality of data is most important."

Jeff Seibert differentiates between the importance of data size for base LLMs and the significance of data quality for fine-tuning purposes.

• Startups with thin wrappers around LLMs are most vulnerable.
• OpenAI may evolve into an infrastructure company like AWS.
• Startups should avoid incremental improvements that are on OpenAI's roadmap.

"If you're working on a use case that's pretty horizontal, that OpenAI is going to need to solve within five years in order to scale, that's not a great investment, and that's not a good use of your time as a founder."

Jeff Seibert discusses the risks for startups in the AI space and the potential direction of OpenAI's business model.## Enterprise Adoption of AI

• Enterprises initially skeptical about cloud services now embrace them due to the expertise of cloud providers in data center operations.
• Enterprises should adopt AI strategically, focusing on customer needs, market, and product development, rather than as a panacea.
• AI should be viewed as a tool to enhance products and services rather than a standalone solution.
• The risk of enterprises adopting AI as a trend without a clear purpose, much like the unnecessary hype around blockchain technology.

"I think this will go in the same direction. You'll have very clear guidelines around how the companies use the data for model training and it's off limits and so on, and sort of that trust will be overcome."

This quote emphasizes that enterprises will eventually trust AI as they did with cloud services, provided there are clear guidelines on data usage for model training.

"Again, to me, focus on your customer, your market, your product need, and view this as a tool, not a panacea, and adopt it strategically on what makes sense and where it's going to push the product forward."

Jeff Seibert suggests enterprises should prioritize their core business needs when adopting AI, using it as a tool to advance their products rather than jumping on the AI bandwagon without a clear strategy.

• AI implementation services will likely become a large market, though not particularly interesting to Jeff Seibert.
• The potential for new, innovative approaches to dominate quickly due to low barriers to entry, as opposed to traditional consulting services.
• Opportunities for startups to disrupt industries with AI, specifically in areas like accounting, rather than expecting established companies to lead the change.

"I think you're probably right in that it's a large market. To me, it's not a very interesting market."

Jeff Seibert acknowledges the potential size of the AI implementation services market but expresses a lack of interest in it due to the likelihood of subpar products and minimal impact.

"Let's build real workflow automation for real people in massive industries that are outdated."

Seibert advocates for the development of practical AI applications that automate workflows and bring significant improvements to outdated industries.

• AI adoption may be faster than previous technological transitions due to the absence of new hardware requirements and familiar UX patterns.
• The ease of online access and the intuitive nature of AI interfaces, like chatbots, could lead to rapid industry disruption.

"It'll be faster for a couple of reasons. So if you look back at mobile, so the iPhone came out in 2007, they opened up the App Store in 2009. By 2011 to 2012, a lot of people were using in building apps and then enterprise adoption even lagged from there. So call it five to seven years. With AI, there's no new hardware to buy, so you don't need this huge purchase price."

Jeff Seibert explains that AI adoption is likely to be quicker than mobile adoption because it doesn't require the purchase of new hardware, which lowers barriers to entry.

• Google may need to aggressively invest in AI to avoid being disrupted by it, potentially cannibalizing their primary revenue source, search.
• The historical precedent of Apple actively disrupting their own products, like the iPod Nano, in favor of better technology.

"They need to go all in on it. I don't think they have a choice. I agree with you. I think it's existential for them because if AI replaces search, their golden goose has been killed."

Jeff Seibert argues that Google must fully embrace AI development to avoid the risk of AI replacing their main revenue generator, search.

• AI models are memory and CPU bound, requiring advancements in memory bandwidth alongside processing speed.
• The cost of compute and queries is expected to decrease rapidly due to technological advancements.
• AI pricing models are likely to follow industry norms, whether per-seat or consumption-based, as AI is seen as a tool rather than a product.

"I would bet on the pace of technology, it will get a lot faster and a lot cheaper. Very, very fast."

Jeff Seibert predicts a rapid decrease in the cost of compute and queries as technological advancements continue at a fast pace.

"So this may be just me, but I very much see AI as a tool, not a product."

Seibert views AI as a technology that will be integrated into existing pricing models rather than altering them.

• Acquiring high-quality, clean data is extremely challenging and has become a proprietary defense mechanism for companies.
• Companies are increasingly restricting access to their data, reversing the trend of openness and API accessibility.
• The importance of proprietary data sets, such as Digits' collection of financial transactions, for training AI models.

"It is extremely challenging. And what's interesting is the counterreaction, because you're seeing Reddit, Twitter, et cetera, shut off APIs, put in more strict rate limits, et cetera, et cetera."

Jeff Seibert notes the difficulty of obtaining high-quality data as companies become more protective and restrict access through APIs and rate limits.

"We have a proprietary data set of 100 million financial transactions, and that's what we can train on and make sure our finance and bookkeeping ais know what they're doing."

Seibert highlights the value of proprietary data, specifically mentioning Digits' financial transaction data set used for training AI models in finance and bookkeeping.

• Google is considered vulnerable due to its reliance on search revenue and slow reaction to AI developments.
• Amazon may need to make strategic acquisitions in AI to strengthen AWS in light of competitors' partnerships.
• Skepticism about the effectiveness of AI features in products where they may not be necessary, such as Dropbox's AI feature.
• Many startups and scale-ups are introducing AI products to follow trends rather than for substantial product improvements.

"I think Google's by far the most vulnerable because again, their business model is pretty binary."

Jeff Seibert identifies Google as particularly at risk in the AI transition, given their heavy reliance on search revenue and slow adoption of AI.

"That is super interesting. I would be very nervous about the OpenAI Azure partnership and of course meta just made a big deal about partnering with Microsoft as well."

Seibert suggests that Amazon should consider acquiring AI companies to compete with the partnerships between big tech firms and AI organizations.

"They really are. And one of the funnier ones to me is Dropbox. Dropbox has a built in AI and I don't know, I mean, I'm sorry, drew, but I just want Dropbox to store the files."

Seibert criticizes the trend of companies adding AI features that may not be necessary, using Dropbox as an example.

• The high failure rate of startups and the importance of discipline in investment amounts and terms.
• The majority of returns in angel investing are on paper, with actual cash returns being rare.
• The unpredictability of startup success, even with seemingly strong founders, markets, and execution.
• The value of a diversified portfolio in angel investing and the pitfalls of overconfidence in individual investments.

"So far, 30 have failed outright, about a third. Another 19 are still at one x, so basically haven't gone anywhere. And if you look at the overall portfolio, really, it's like ten of them matter."

Jeff Seibert shares his personal angel investing data, highlighting the failure rate and the concentration of value in a small number of investments.

"No matter how great you think the company is, how great you think the market is, how great you think the founder is, it is still damn hard."

Seibert reflects on the inherent difficulty of achieving success in startups, regardless of the apparent strengths of the company or founder.## Traction and Sustainability

• Traction in consumer sectors may not indicate long-term sustainability.
• Real traction in enterprise software could be more sustainable.
• Rapid growth (e.g., reaching 10 million ARR in 18 months) is not indicative of future, larger-scale success.

"Traction doesn't mean sustainable."

"On the consumer side, I'd say real traction on enterprise software. Okay. Potentially more sustainable."

These quotes emphasize the difference between initial traction and sustainable growth, particularly contrasting consumer sectors with enterprise software.

• Jeff Seibert invested in a friend's startup, which pivoted multiple times before succeeding.
• The key takeaway is the importance of a founder's grit and mentality over the initial idea.
• Outlier successes can significantly outperform other investments.
• Timing and early entry at low valuations are critical for high returns.

"Nothing. It's all about a founder's grit and mentality, and you can't back the early idea and think that that's going to be it."

"It's remarkable how the outliers outclass the rest."

Jeff Seibert discusses his successful investment and the lessons learned about the importance of the founder's qualities and the impact of outlier investments.

• Market timing can significantly affect investment returns.
• Partial exits can be strategic, allowing investors to secure returns even if future valuations are uncertain.
• The decision to sell shares partially is influenced by concerns over inflated valuations.

"Yes. And getting in very early at a very low valuation. That's another great lesson learned for folks, actually."

"I'm worried that valuation is fake. I don't think it'll ever return the remainder."

Jeff Seibert reflects on the benefits of entering investments early and the strategy behind taking partial exits when valuations seem unsustainable.

• The appropriateness of investors asking for cash back from founders is situational.
• If the founder and team display commitment and grit, investors should continue to support them.
• In cases of wavering founders or poor performance, it may be reasonable to consider cash back.

"So if the founder is still excited, the team is there, they have grit. I don't think you can ask for cash back."

"But in general, I would bias towards supporting the founder."

Jeff Seibert shares his nuanced view on when it might be acceptable for investors to request cash back, emphasizing founder support.

• Economic downturns can influence talent migration within the tech industry.
• Employees may opt for safety and stay at overvalued companies due to higher salaries.
• Down markets can concentrate talent in fewer ventures, increasing talent density.

"A lot of folks, you're right, are trapped in companies with high valuations where their options are likely underwater."

"In down markets you can build higher talent density because it's harder to raise your own money."

Jeff Seibert talks about the challenges and opportunities for talent migration in different market conditions, highlighting the potential benefits of talent concentration in down markets.

• Successful founders are not always the most technically skilled but often deeply understand their market.
• Customer obsession and a personal connection to the problem space can drive success.
• An interesting pattern observed is that successful founders often moved a lot in their early childhood.

"It's not even necessarily the most techie founders that are successful, but the most sort of customer obsessed."

"They've just felt this personal identity come to bear on what they're doing."

Jeff Seibert discusses the traits that can contribute to a founder's success, emphasizing customer focus and personal investment in the business.

• Jeff Seibert believes AI will enhance productivity rather than replace jobs.
• Historical fears of new technologies eliminating jobs have been unfounded.
• AI should be viewed as a tool that enables more efficient work, similar to past technological advancements.

"So I would say, actually AI won't replace many jobs. I think actually it'll drive productivity, not replacement."

"People hated personal computers in the 70s because they thought it was going to take over and replace them."

These quotes reflect Jeff Seibert's contrarian view on AI, suggesting it will augment rather than replace human labor, drawing parallels with past technological evolutions.

• Startups should concentrate on understanding and serving their customers rather than on competition.
• Focusing on customers can guide product development and business direction more effectively than monitoring competitors.

"Ignore them completely. Focus on the customer."

"As a startup, the market is way larger than you could possibly capture this year."

Jeff Seibert advises startups to prioritize customer needs over competitor actions, arguing that this focus is more conducive to growth and success.

• Quick decision-making within 24 hours is crucial for maintaining momentum.
• Most decisions are reversible and should not impede progress.
• Moving in any direction is better than stalling due to indecision.

"When your team comes to you with a question, you need a decisive answer within 24 hours."

"If you're not moving, you're not learning."

Jeff Seibert stresses the importance of rapid decision-making in startups, linking movement and learning to progress and adaptation.

• Jeff Seibert predicts runaway climate change within the next decade.
• The global response to climate change is inadequate and lacks prioritization.
• The impact of climate change is unevenly distributed, posing challenges for international cooperation and emerging economies.

"Runaway climate change is less than ten years out."

"It's still shocking to me that this is not like the singular top priority in any political campaign in the entire world."

These quotes convey Jeff Seibert's concern about the imminent threat of climate change and the lack of political urgency to address it.