finance
The Twenty Minute VC (20VC) interviews the world's greatest venture capitalists with prior guests including Sequoia's Doug Leone and Benchmark's Bill Gurley. Once per week, 20VC Host, Harry Stebbings is also joined by one of the great founders of our time with prior founder episodes from Spotify's Daniel Ek, Linkedin's Reid Hoffman, and Snowflake's Frank Slootman. If you would like to see more of The Twenty Minute VC (20VC), head to www.20vc.com for more information on the podcast, show notes, resources and more.
Nvidia CEO Jensen Huang has boldly declared that Artificial General Intelligence (AGI) has now arrived, marking a significant shift in the perception and development of AI. This pronouncement suggests a potential acceleration in the AI model race.
OpenAI has announced the release of its latest AI model, GPT Astra. This development is seen as a direct contribution to the intensifying race among major players in the AI model space, with products like GPT Astra potentially setting new benchmarks.
Index Funds have reportedly ceased investing in a company referred to as 'Town' following its involvement with 'Instinct'. The exact nature of this involvement and the implications for Index Funds' investment strategy remain unclear.
The competitive landscape for AI assistants continues to evolve with Matter announcing the release of its product. This move occurs amidst a broader, accelerated race in the development and deployment of AI assistant technologies.
Wonderful has achieved a significant financial milestone, doubling its valuation to $5 billion in just six months. The company also successfully raised $170 million through secondary transactions.
A debate is emerging regarding the ethics and sustainability of AI startups that operate by 'breaking rules,' whether legal or terms of service. While some see this as a necessary path to innovation, others caution against the long-term viability of such models.
The discussion highlights the critical role of form factor and user integration in the success of AI products. It's suggested that products seamlessly embedded into existing platforms, like WhatsApp, have a significant advantage.
Venture capitalist Jason Calacanis voiced strong doubts about the business model and long-term prospects of the company 'Instinct'. He predicts a potential downfall due to uninteresting technology, weak moats, and the impending wave of open-source competitors.
Jean Denis (JD), founder of Town, discussed the company's pivot from AI tax preparation to an AI assistant that integrates with email and calendar. The initial venture into tax prep did not achieve sufficient product-market fit, leading to a strategic reset and focus on areas with high user engagement like email and calendar.
JD, founder of Town, believes that AI assistants are a top-three priority for major tech companies like Google and Apple within the next 12 months. He acknowledges the competitive landscape but feels confident in his company's direction.
JD argues that the winning AI assistant product will likely have a network effect at the agent level, referencing Town's 'agent to agent' feature. He also highlights distribution as a critical factor, with Meta and WhatsApp being significant potential competitors due to their existing user base.
JD posits that AI is fundamentally changing the software landscape, predicting that AI assistants will become the primary entry point for digital interactions, eventually replacing traditional websites and mobile apps. He believes owning the user's primary AI interface will be crucial for market dominance.
Town's AI assistant features an 'agent to agent' functionality where one assistant can query another if it doesn't have the answer. This feature is highlighted as a key aspect of Town's strategy to build a network effect, making it difficult for users to switch to competing products once integrated within a team.
JD discusses the perception of the AI assistant market, noting that while some see it as commoditized, he views talking about 'moats' as a luxury. He believes success will come from deep product-market fit and attracting a large user base, rather than solely relying on defensibility at this stage.
Cliff Weitzman, CEO of Speechify, discussed the company's strategy of purchasing Nvidia GPUs, even at a premium of $100,000 per unit for early access. He explained that this investment is driven by the economics of training large language models and the need for dedicated, high-memory infrastructure that is not readily available through cloud providers.
Cliff Weitzman stated that Speechify's latest model, Speechify 3.2, is ranked number one globally for quality, surpassing even frontier labs. He also highlighted its cost-effectiveness, being 10 times more affordable than solutions like ElevenLabs.
Cliff Weitzman presented a financial argument for purchasing GPUs over renting them. He calculated that renting an H100 GPU for a year could cost between $35,000 to $50,000, whereas buying one for $30,000 and using it for its warranted three-year lifespan (or longer) is more economical.
Cliff Weitzman elaborated on why Speechify chooses to own its GPU infrastructure, emphasizing the necessity of co-located memory with a large cluster of GPUs for effective large-scale training. He stated that renting from cloud providers doesn't offer the required memory accessibility for their specific training needs.
Cliff Weitzman argued that running open-source models on self-owned hardware is significantly more cost-effective than using third-party token-based services. He noted that this approach allows for a cost per token that is a 'fraction of a fraction of a cent'.
Cliff Weitzman shared an observation that when Speechify engineers used rented GPUs, they tended to be 'parsimonious' due to the high cost. This led to the idea of owning GPUs to encourage more liberal usage for model training, comparing it to Michael Jordan needing affordable access to a basketball court.
NVIDIA has announced a record-breaking quarter with $96.2 billion in revenue, exceeding analyst expectations. The company is also nearing a $12.9 billion deal to acquire Hugging Face, a move that could significantly impact the AI landscape.
NVIDIA has provided a strong outlook, projecting 70% revenue growth for the next fiscal year, significantly higher than the 44% expected by analysts. This projection is driven by sustained, intense demand for compute power.
The acquisition of Hugging Face by NVIDIA for $12.9 billion is seen as a strategic move to win across all market segments, including open-source AI. This positions NVIDIA to capture a significant portion of the AI dollar, even if it means subsidizing open-weight models.
OpenAI has reportedly cut off access for Cursor, an AI-powered code editor. This action drew a notable response from Mike Trout, who expressed concern over the impact on traffic, and a brief comment from Elon Musk regarding Sam Altman.
Discussions arose regarding Cursor's flexibility in using custom codecs and the perceived inconsistency in data related to its usage over the past 60 days. The reliability of the '5% traffic' figure for OpenAI was questioned.
Instinct, a new AI assistant from Silicon Valley, has reportedly raised capital at a $2.5 billion valuation. The company is generating significant excitement within the tech community.
Aaron Katz, founder of ClickHouse, notes that the current AI growth cycle is accelerating at an unprecedented pace compared to previous internet and mobile cycles. He highlights that companies are growing rapidly and the demands on their systems are unlike anything seen before.
Aaron Katz suggests that while many AI companies operate with lower gross margins than traditional SaaS companies, investors may be satisfied as long as a path to margin expansion is demonstrated. He contrasts this with concerns about gross margins five years ago in enterprise software.
Aaron Katz identifies the durability of revenue as the single biggest risk for AI companies, particularly generative applications. He notes that switching costs can be very low for these applications as model providers rapidly leapfrog each other.
Aaron Katz describes a moment of 'AI awakening' at ClickHouse, where they realized they were not adopting AI coding applications fast enough. After sending an email to Anthropic, the company saw explosive growth in their usage of these applications.
Aaron Katz reflects that if he could change one thing over the past two years at ClickHouse, it would be increasing sales capacity earlier. He contrasts their 100 sales people with competitors who have thousands, covering fewer use cases.
Aaron Katz shares his biggest takeaway from his 12 years at Salesforce, working with Mark Benioff: 'That you can overestimate what you can achieve in one year and underestimate what you can achieve in five.' He notes how Benioff's vision and marketing created a perception that allowed Salesforce to grow into a major competitor.
Aaron Katz states that AI tooling has enabled ClickHouse to ship products faster and enter new product categories two years ahead of schedule. He explains their acquisition strategy: building organically when internal teams can innovate, and acquiring when founders are building on ClickHouse in strategic future categories.
ClickHouse has acquired Langfuse, a Berlin-based company with three founders, to bolster its capabilities in agent observability. Aaron Katz believes this technology will be essential for every enterprise.
Aaron Katz prioritizes revenue growth for his company, stating he worries less about expenses like coding agents if sustained revenue growth is observed. He believes that covering a broad service area and accelerating the roadmap justifies the spending.
Aaron Katz draws parallels between ClickHouse's sales approach and those of DataDog and Snowflake. He initially favored DataDog's product-led growth (PLG) but now recognizes the need to layer in an enterprise sales motion, learning from his Salesforce experience.
NVIDIA is making significant moves in the AI space, acquiring Poolside for $12 billion and investing $1 billion in Mercor and Perplexity's latest funding rounds. The Poolside acquisition includes licensing their model factory for $6 billion and moving 109 engineers to NVIDIA's Nemo Tron team. This strategic expansion aims to bolster NVIDIA's AI ecosystem and full-stack capabilities.
OpenAI's CFO, Sarah Fry, has confirmed that the company is definitely going public this year. This announcement comes as Anthropic is also pursuing an IPO, indicating a significant shift for major AI players into the public markets.
An investor letter revealed that a company was unable to raise $2 billion for GPU acquisition, forcing them to accept a $6 billion acquisition by NVIDIA. The letter detailed the immense capital intensity required for AI model development, suggesting that competing at the frontier is becoming increasingly difficult for many.
The conversation touches on the implications of NVIDIA's acquisitions and investments, suggesting a trend towards consolidation in the AI infrastructure market. While NVIDIA's strategy benefits existing investors and founders, questions remain about the survival prospects for smaller AI startups.
The discussion highlights that despite challenges in raising capital, companies can still achieve significant value through acquisitions in a rapidly growing market. The focus shifts from standalone profitability to the cash extractable from acquirers who possess capital, GPU access, and market position.
Matt Swulinski, who has experience with Superhuman and Whisperflow, argues that the e-commerce playbook is the correct model for SaaS growth. He emphasizes the importance of paid advertising and user-generated content (UGC) for scaling companies.
Matt Swulinski explains that Meta's 'Andromeda' update means the creative itself serves as the targeting mechanism. He advises startups to invest heavily in creative for paid channels, especially in the early stages.
Matt Swulinski discusses the challenge of producing 400-500 new creatives per month for paid channels. He suggests starting with a creator program, where a percentage of ad spend is given to individuals to create ads, and gradually scaling to agencies and in-house teams.
Matt Swulinski highlights a significant gap in the SaaS industry: the lack of out-of-the-box solutions for attribution and conversion tracking, unlike in e-commerce. He notes that SaaS companies often build these systems in-house, which can be complex.
Matt Swulinski advises that when starting paid acquisition for a SaaS business, the primary focus should be on optimizing for the most critical event, such as a subscription or a download. He stresses the importance of accurate conversion tracking for effective campaign optimization.
Matt Swulinski discusses how Product-Led Growth (PLG) principles still apply in the age of AI agents. He believes that optimizing for how agents make decisions and select tools is crucial, and companies strong in PLG will be better positioned for this agentic layer.
Matt Swulinski reflects on Superhuman's growth strategy, noting its initial reliance on founder-led referrals. He explains the transition to a paid advertising model as the company hit an asymptote with organic growth, adopting an e-commerce playbook.
Canva has reportedly cut its 2026 growth targets by one-third, citing the escalating costs associated with its AI features. CEO Melanie Perkins indicated that while growth is still projected at 20%, the increased expenses from utilizing advanced AI models are impacting their earlier projections. The company is reportedly developing its own in-house AI models to mitigate these costs.
Jeff Dean, a prominent figure in AI and a long-time Google employee, has left the company after 27 years. His departure is noted in the context of broader concerns about talent retention at Google, with Demis Hassabis also reportedly stepping back from his role at Google DeepMind. The discussion highlights the significant impact such departures can have on a company's valuation.
The discussion delves into the impact of AI on creative software companies like Adobe, Figma, and Canva. Jason argues that while companies like Canva are experiencing growth slowdowns due to AI serving costs, the larger question is whether these companies can integrate AI as a feature or become obsolete. He notes that AI is making many 'no-code' tools redundant.
Elon Musk has announced a new, highly ambitious project called TerraFab. The project aims to secure the first real installment of $16.8 billion and could structurally alter Musk's reliance on TSMC for manufacturing.
The podcast segment emphasizes the growing importance of AI risk management in business. With every new AI tool and integration, there's an increased opportunity for things to go wrong, and existing security programs may not be equipped for AI's rapid growth. Vanta is presented as a solution for monitoring and managing these risks.
The podcast highlights Deal.com as a solution for startups looking to scale faster by hiring globally. Deal handles the complexities of international employment, including payroll, visas, onboarding, HR, IT, EOR, benefits, and compliance, allowing founders to focus on growth.
Frame.com is promoted as a pro AI website builder that bridges the gap between AI-generated ideas and production-ready websites. It allows for on-canvas editing, custom code components, SEO optimization, and enterprise-level features like premium hosting and security.
The discussion posits that the era of 'no-code' tools, which allowed humans to build without extensive coding knowledge, is ending due to AI. Specifically, agentic AI is becoming sophisticated enough to bypass traditional tools like Canva and AirTable, with consumers increasingly using AI platforms like ChatGPT for tasks previously handled by specialized software.
Airtable, a prominent company from the last decade, has been acquired by the European-based Bending Spoons for $1.285 billion. The company had reported $485 million in ARR with 20% year-on-year growth, but the acquisition price is significantly lower than its previous $11 billion valuation.
The podcast transcript mentions Moonshot AI's significant funding round, raising $3.5 billion at a $35 billion valuation. This is framed within a broader discussion of market dislocations and the increasing importance of compute as a priced asset.
The transcript highlights a security incident where Anthropic's model reportedly breached the security of three companies. This event is presented as a continuation of ongoing security problems in the AI space.
Palantir is noted for beating other Big Tech companies in earnings, attributed to their significant sales of compute. The discussion contrasts this success with the struggles of other companies in the market.
The episode discusses Leo Achenbrenner's 'situational awareness imploding,' linking it to issues in portfolio construction and a speculative investment strategy that led to significant losses. The speaker mentions losing all their money in a week due to unhedged positions.
A bold prediction is made that all consumer applications will undergo a rewrite within the next five to ten years, driven by the pervasive influence and advancements in AI. This is presented as a consequence of 'insatiable demand' for AI capabilities.
The SaaS marketplace is facing uncertainty, with discussions centered on whether the current situation represents a temporary pricing dislocation or a fundamental shift in expected long-term growth rates for SaaS companies. This uncertainty is impacting valuations and investment decisions.
Nikasha Aurora suggests 'founder fatigue' may have played a role in the Airtable acquisition, even though the company demonstrated strong revenue figures like $450 million in ARR and 20% growth. She notes that despite meeting many criteria for an attractive target, fewer potential buyers than expected stepped up.
Rory highlights Bending Spoons' strategic advantage in the current market, noting their capital and ability to use a quoted currency to acquire other companies. He suggests this positions them well to take advantage of opportunities to 'hoover up' assets.
A discussion compares Air table's approach to AI infusion with automotive analogies, questioning whether it was a superficial 'Mercedes action' (sprinkling AI) or a fundamental 'Waymo action' (building a self-driving car). The concern is that companies might be adding AI 'pixie dust' without core transformative AI capabilities.
Joon Sung Park, founder and CEO of Semily, shared his thesis that AI companies need defensible data strategies. He discussed Semily's work in simulating human behavior through agents trained on vast amounts of human interaction data, aiming to predict and influence future outcomes, rather than just predict them. Park highlighted the importance of collecting 'real' behavior data, such as from A/B tests and randomized control trials, to understand cause and mechanism.
Semily's early work involved creating a game town populated with 25 NPCs using GPT-3.5, memory, planning, and reflection capabilities. A key experiment during Valentine's Day showed these agents self-organizing, planning parties, and decorating a cafe, demonstrating emergent complex behavior. This project was a foundational step in developing agents that remember interactions and exhibit personality.
Joon Sung Park explained that Semily differentiates itself from other AI companies like OpenAI by focusing on simulating human subjectivity, values, and biases, rather than pure rationality. Semily aims for its models to make similar 'mistakes' and exhibit the same biases as humans. Their data collection prioritizes understanding cause, mechanism, and counterfactuals over simple prediction.
Semily's data collection strategy centers on sourcing 'everyday people' to represent the general population, rather than expert programmers or scientists. The company asks individuals to share life stories, origins, and problem-solving experiences to gather nuanced behavioral data. This approach aims to ensure representativeness and capture the core nature of human decision-making.
Travis Kalanick has re-entered the venture capital scene, reportedly raising $1.7 billion for his new venture, Atoms. The exact focus of Atoms remains somewhat vague, but it's positioned as a company that will significantly impact the tech landscape.
NVIDIA CEO Jensen Huang has broken his silence on X regarding open-weight models, noting that major players like Anthropic and Dario have not signed onto a 50-company open-weight slate. He also expressed concerns about AI risk, suggesting that open-weight models should be kept out of the US.
Google Cloud has demonstrated strong acceleration, growing by 82%. However, this growth was accompanied by the company posting its first-ever negative free cash flow, leading to a market downturn.
Francisco Partners has successfully raised $21 billion. This significant capital raise positions the firm to make substantial investments in the technology sector.
Etched, an educational technology startup, has secured $300 million in funding. The company aims to challenge Nvidia's dominance in the market.
The speaker predicts that within the next 24 months, every company will experience a security breach due to LLM agents, noting that many may already have occurred but are not being disclosed. This highlights a significant and growing AI risk.
Osvald Nitzky, CPO at McCord, stated that open source model improvements do not cannibalize their core business. He explained that data is most valuable at the frontier of model performance, and open models simply raise the baseline of what customers are interested in, allowing McCord to continue growing by serving new capabilities.
Osvald Nitzky expressed skepticism about the claim that 90% of enterprise workflows can be handled by open models, suggesting these calculations might overlook latent demand for complex tasks. He cited long-horizon tasks like automating procurement teams as examples of use cases not yet captured.
Osvald Nitzky noted that large enterprises show skepticism and fear towards working with frontier models, particularly for core business workflows that differentiate them. However, for more general tasks like HR or procurement, companies are more open to using proprietary models.
Osvald Nitzky believes that approximately 50% of long-horizon workflows are currently addressed by top AI models, based on their Apex benchmarks. He also noted that for sufficiency-based workflows, like updating a CRM, performance ceilings are quickly reached, while areas like legal or medical advice offer continuous improvement potential.
Osvald Nitzky agrees with the notion that specialized models for each company will be the future, as it aligns with McCord's strategy of providing enterprise-specific evaluation and training data. He believes the market diversity for this depends on the demonstrable value customers receive from these specialized models.
Osvald Nitzky believes that companies are not currently facing an ROI problem with AI spend but are in a phase of exploration and experimentation. He suggests there is more tolerance for calculating ROI correctly, and while some spending is tightening, the rapid pace of AI development means ROI calculations can shift significantly.
Osvald Nitzky advised that optimizing AI spend depends entirely on the use case, particularly distinguishing between spending for growth and spending for less critical functions. He noted that for hypergrowth companies like McCord, high spending on AI is justified by insatiable customer demand and the need to scale rapidly.
Osvald Nitzky believes the current AI paradigm makes product management more challenging, requiring a focus on simplifying the product service area and identifying scalable workflows. He notes that with engineering less of a bottleneck, understanding user needs and identifying revenue-driving features has become the new bottleneck.
Apple has filed a lawsuit against OpenAI, alleging trade secret theft. The suit names Tani Tang, a senior executive at OpenAI and a former Apple veteran, and Chang Lu, a six-year Apple employee. The core allegation is that Lu took proprietary information from Apple to OpenAI, with Tang allegedly encouraging this action.
Meta is actively competing in the coding model space, with "Zach" being mentioned as back on X. The discussion implies Meta is looking to leverage its position against competitors like Codex and Claude.
SK Hynix has priced its Nasdaq listing at $26.5 billion, bringing its compute and infrastructure layer to the public domain. This move is seen as significant for the sector.
The accelerating pace of AI adoption introduces new risks, and existing security programs may not be equipped to handle them. Vanta is presented as a solution to monitor and mitigate these risks across various platforms and AI tools.
Deel aims to help founders scale startups faster by handling the complexities of global hiring, including payroll, visas, onboarding, HR, IT, EOR, benefits, and compliance. The service is used by over 40,000 companies.
Framer is highlighted as a pro AI website builder for creators and businesses focused on detail. Its AI agents help bridge the gap between AI-generated ideas and production-ready websites, offering features like custom code components and SEO optimization.
The discussion delves into the legal ramifications of Apple's lawsuit against OpenAI, focusing on trade secret theft. It highlights the risks for individuals involved, such as Chang Lu and Tani Tang, and the broader implications for employee mobility and corporate espionage in the tech industry, referencing California's employment laws.
Wix CEO Avishai Abrahami believes the public market undervalues his company, stating that its current market cap of $2.8 billion does not reflect the value of its Base44 product alone, which he estimates should be worth $8 billion. He argues that while peers in 'vibe coding' receive higher multiples, Wix is not being adequately recognized for its growth and product offerings.
Wix CEO Avishai Abrahami shared his unconventional sleep schedule, typically going to sleep between 5-6 AM and waking around 11 AM, which he finds beneficial for focused work. He contrasts this with the perceived necessity of early sleep schedules in some tech circles, noting that wearable tech companies like Aura have adapted their sleep profiles to include 'very late sleepers'.
Wix CEO Avishai Abrahami asserted that Salesforce's primary value proposition is the trust it instills in large companies to handle their sensitive customer data, rather than its CRM functionality. He believes this trust makes it incredibly difficult for competitors to displace Salesforce, even with advancements in other areas like 'cloud code projects'.
Wix CEO Avishai Abrahami described the company's recent buyback program as a 'disaster', comparing its success to the Titanic. He feels the public market is not properly valuing Wix, particularly its Base44 product, leading to a disconnect between the company's performance and its market capitalization.
Arvind Jain, co-founder of Glean, believes that enterprise leaders are both terrified of and dependent on frontier AI model providers. He acknowledges a skepticism towards these providers due to concerns about business continuity and over-reliance on technology, but also notes that many enterprises are exploring AI despite these fears.
Arvind Jain, co-founder of Glean, explained that the company started as an enterprise search engine to help employees find information across various systems. As AI models improved, Glean evolved into an AI platform, now functioning as a 'super set' of tools like ChatGPT, connecting with a company's internal context to act as a coworker.
Arvind Jain stated that the main reason enterprises are considering open-source AI models is the prohibitive cost of using frontier model providers. Many companies are finding that their AI budgets are exhausted within months due to the expense of these advanced models, making open-source alternatives increasingly attractive.
Arvind Jain highlighted a key concern for enterprises: the potential for operational dependence on frontier AI model providers. As AI agents handle more work, companies worry about losing control over their data and the accumulated institutional learning, which could become embedded within these external models.
Arvind Jain views open-source AI models as both a threat and an opportunity. He acknowledges that they pose a challenge to incumbents like OpenAI and Google, but also notes that the rapid improvement and accessibility of open-source models are driving innovation and allowing startups to build differentiated products on top of them.
Arvind Jain describes his entrepreneurial drive as stemming from a fear of losing rather than a thrill of winning. He believes this constant worry about what could go wrong is what fuels his persistence and success, even after achieving significant milestones like the IPO of Rubric.
Showing the latest 15 of 20 covered episodes.