Question 1

How can startups use large language models to improve their products?

Accepted Answer

_Simplify interfaces and gain user insights through language_

As a startup founder, think about which interface will be changed by language input. Whenever users needed to use their mouse or learn complex clicking and dragging, you suddenly have the ability to let them do it via language. This simplifies the interface dramatically. Additionally, language prompting gives you more feedback. Before, you'd create an interface and see how users clicked buttons and moved around without really knowing their intentions. Now with language interfaces, you actually see what they're searching for and asking for. This gives you valuable input for building your product.

— [AI as an interface](https://www.lutzfinger.com/transcript/2023-04-24-ai-as-an-interface/) · _The Edge_

Question 2

How could Europe break down existing monopolies to help AI startups compete?

Accepted Answer

_Portability and open weights could level the AI playing field._

Europe must find ways to level the playing field by breaking down legacy moats. Let users port their social connections or AI history from one platform to another. Ironically, European privacy laws, meant to protect consumers, often reinforce monopolies instead. Additionally, what if Europe required AI companies to make their weights open? This would give European startups a fighting chance to build on shared infrastructure instead of starting from scratch. Europe's push for data spaces is well-intentioned but overengineered. Data is AI's oxygen, and enabling true data access is essential for competition.

— [Nvidia Wants To Power Europe's AI Dreams. The EU Isn't Ready](https://www.forbes.com/sites/lutzfinger/2025/06/16/nvidia-wants-to-power-europes-ai-dreams-the-eu-isnt-ready/) · _Forbes_

Question 3

How do you decide if AI is the right fit for your product problem?

Accepted Answer

_Not every problem needs AI; start with whether complexity demands it._

We created an AI fit analysis with portfolio companies. It starts with questions around precision and output quality, focusing on data quality. Then ask: Do you benefit from modeling complex decisions versus just a few? If it's a linear problem, a rule engine works. But if you have a complex decision space or nonlinear problems like search engines or recommendation engines, then it's time to deploy AI to find tiny clues. This doesn't always mean large language models. Sometimes simple regression analysis is good enough and more precise.

— [Building AI products: 5 lessons from our founders' workshop](https://www.lutzfinger.com/transcript/2024-08-01-building-ai-products-5-lessons-from-our-founders-workshop/) · _The Edge_

Question 4

How does Lutz compare Clubhouse to the early internet and talk radio?

Accepted Answer

_Clubhouse mirrors talk radio but doesn't exploit digital advantages._

Clubhouse is Internet 1.0 of Talk Radio. Ten years before the internet browser was invented, Talk Radio became popular for similar reasons: ease in distribution and proliferation of content. AM Bands allowed for more bandwidth to distribute content. Four decades later, the internet brought the same two effects. Clubhouse follows the rules of Internet 1.0, meaning it is not using the properties of digital to their full advantage. This is why it faces challenges in helping users find content and enable engagement.

— [Clubhouse's Future Depends On Data - How To Build A TikTok Like Algorithm](https://www.forbes.com/sites/lutzfinger/2021/02/09/clubhouses-future-depends-on-datahow-to-build-a-tiktok-like-algorithm/) · _Forbes_

Question 5

How does Neuralens measure brain health non-invasively compared to current methods?

Accepted Answer

_Replacing skull drilling with light-based sensors for brain monitoring._

Currently, doctors literally drill a burr hole through your skull with a household drill and insert sensors through brain tissue to measure oxygen saturation, blood flow, and intracranial pressure. This invasive surgery is only done on 200,000 to 300,000 patients yearly, though over 10 million neurocritical care patients could benefit. Neuralens uses advanced methods of measuring how light interferes with brain tissue to quantify blood flow, blood volume, and intracranial pressure non-invasively. We can stick sensors on the head or shine a laser from 30 centimeters away to make these measurements.

— [Founder Coaching: Tips for your AI Journey](https://www.lutzfinger.com/transcript/2024-07-26-founder-coaching-tips-for-your-ai-journey/) · _The Edge_

Question 6

How does your company Gooder AI address the gap between technical metrics and business value?

Accepted Answer

_*Translating predictive model performance into profit and savings estimates*_

Gooder addresses one particular place where there's a need for a new category of software: the metrics. It's a business user console for understanding potential value and making decisions about deployment. Typically, data scientists present technical metrics like area under the curve, precision, or recall, but very rarely go to stakeholders and say if you deploy this model, it has potential value of this much profit or savings. We provide visualization where both dimensions are business relevant, like the decision threshold and business metrics like savings or profit, so stakeholders can see exactly what their options are.

— [The Value Translation Gap: AI's Deployment Problem](https://www.lutzfinger.com/transcript/2024-11-28-the-value-translation-gap-ais-deployment-problem/) · _The Edge_

Question 7

How does Zenith OS reduce costs in neuropsychiatry treatment?

Accepted Answer

_AI cuts neuropsychiatry costs by reducing required therapist headcount per session._

Neuropsychiatry treatments like ketamine and psilocybin for PTSD, depression, and anxiety are 30 to 40 percent more effective than SSRIs but remain inaccessible due to high human capital costs. Currently, two therapists must be present during treatment because patients are in a vulnerable state and there are sexual harassment concerns. Each therapist charges $100 to $150 per hour. Zenith OS uses AI to augment administrative tasks and help with patient recommendations, allowing clinics to operate with one therapist instead of two, significantly reducing costs and expanding access to these groundbreaking treatments.

— [Founder Coaching: Tips for your AI Journey](https://www.lutzfinger.com/transcript/2024-07-26-founder-coaching-tips-for-your-ai-journey/) · _The Edge_

Question 8

How much capital was invested in the digital health startup landscape in 2021?

Accepted Answer

_Venture capitalists invested $29.1B in digital health startups in 2021._

In 2021 alone, the digital health startup landscape saw $29.1B of capital being invested by Venture Capitalists, and many of these startups are adding LLMs to their products. The landscape is exploding. At my day job as President of Product and Technology at Marpai Inc., our core aim is to help self-insured companies not only administer their claims but also help members find the best provider for their needs.

— [Large Language Models & AI In Healthcare](https://www.forbes.com/sites/lutzfinger/2023/03/27/large-language-models--ai-in-healthcare/) · _Forbes_

Question 9

If OpenAI doesn't have a model moat, what kind of moat does it actually have?

Accepted Answer

_User base of 100M offers paths to sustainable competitive advantage_

OpenAI doesn't have a model moat, but it has a 100 million users moat. ChatGPT currently has over 100 million users, which creates opportunities to build sustainable advantages. There's much that can be done to create lasting competitive protection: User Experience, Data Access, Workflow Integration, LLM governance, and regulatory protection. While the business model should give us pause, the user base provides a foundation for building real defensibility. The moat comes from how you serve and retain those users, not from the underlying technology itself.

— [OpenAI Isn't Going Bankrupt, But It Has A Business Model Problem](https://www.forbes.com/sites/lutzfinger/2023/08/18/is-openai-going-bankrupt-no-but-ai-models-dont-create-moats/) · _Forbes_

Question 10

Is ChatGPT actually at risk of shutting down despite the bankruptcy headlines?

Accepted Answer

_Microsoft backing ensures survival despite user decline and cost concerns_

No. ChatGPT has rapidly evolved into a prominent consumer service. No other application has grown at such a pace, and it's here to stay. Microsoft sees value in OpenAI as a tool to compete with Google and will likely continue funding server costs. While OpenAI has seen a drop in users, bankruptcy is off the table. The attention-grabbing headlines don't reflect the reality. ChatGPT isn't going anywhere. The real issue isn't survival but rather the underlying business model, which has concerning weaknesses around creating sustainable competitive advantages.

— [OpenAI Isn't Going Bankrupt, But It Has A Business Model Problem](https://www.forbes.com/sites/lutzfinger/2023/08/18/is-openai-going-bankrupt-no-but-ai-models-dont-create-moats/) · _Forbes_

Question 11

Should founders focus on building new AI models or developing applications with existing technology?

Accepted Answer

_Applications beat innovation as VCs shift from research to practical tools._

Take the existing capabilities and apply them. Do we really need more innovation, or is it fine to say large language models are very good, you can chain them, combine them with traditional machine learning techniques, and build an application around it? I want more practicality, more tooling around it. The VC world actually sees that the appetite to go on risky, more researchy projects has gone down. VCs want more applications. Founders should focus on taking existing capabilities and applying them rather than pursuing foundational model innovation.

— [AI Gold Rush: Is the Market Overheating?](https://www.lutzfinger.com/transcript/2024-08-08-ai-gold-rush-is-the-market-overheating/) · _The Edge_

Question 12

What advantage do existing platforms like Google, Meta, and TikTok have in the bot space?

Accepted Answer

_Existing user access matters, but cross platform integration creates opportunity_

Companies with existing access to users and data will likely have the majority chunk of the market. While we are still figuring out style, entertainment value, and answer quality, TikTok, Google, Meta, X, and Snap are all putting out bots to monetize their existing access. They already have the right information and the right tools. However, there is an opportunity for cross platform bots that operate across different systems, like accessing both SAP and HubSpot, or working across Snap and LinkedIn, following users regardless of platform.

— [The future of AI bots: winning with personalization, conviction, and cross-platform integration](https://www.lutzfinger.com/transcript/2024-08-22-the-future-of-ai-bots-winning-with-personalization-conviction-and-cross-platform/) · _The Edge_

Question 13

What advice do you have for someone considering building a fintech or launching in Africa?

Accepted Answer

_Hard lessons from exploring Africa and fintech opportunities_

For Africa, most successful businesses are traditional, non-tech businesses like furniture or services. Building something successful requires being hands-on, spending 10 years in the country, and building local relationships. It's really a lifestyle question. For fintech, the space has been extremely hot over the last four or five years. I think you either need to go very close to the rails, like payment rails and banking APIs, or do something very verticalized like payments for healthcare in Europe connecting insurance refunds. Otherwise it's super difficult to create something that works in fintech.

— [AI Products: Why ‘Good Enough’ Beats Perfect](https://www.lutzfinger.com/transcript/2024-11-14-ai-products-why-good-enough-beats-perfect/) · _The Edge_

Question 14

What are the three key takeaways about competitive advantages with LLMs?

Accepted Answer

_Model advantages fade, data and brand loyalty offer lasting protection._

First, be cautious with claims about a Model Moat. A model moat is most likely only transient at best, and models will often be available as open source. Second, focus on Data Moats. Data moats offer a more sustainable advantage, and exclusive access to proprietary data can be a significant barrier to entry, but data needs to create value. Third, Stickiness Moat. Brand can create a powerful competitive advantage if startups focus on user happiness.

— [What Is The Competitive Advantage Of LLMs Like ChatGPT For Your Business? Three Takeaways.](https://www.forbes.com/sites/lutzfinger/2023/04/04/what-is-the-competitive-advantage-of-llms-like-chatgpt-for-your-business-three-takeaways/) · _Forbes_

Question 15

What competitive challenges will OpenAI face in 2025?

Accepted Answer

_Tech giants with data and customers challenge OpenAI's technology bet._

OpenAI will face serious competition in 2025 from Salesforce, Microsoft, Google and Amazon, all of which would like to dominate the AI agent market. These competitors have access to closed datasets and customer bases. OpenAI is betting on its ability to outperform everyone else with superior technology. That's why Altman confidently claims we'll see AGI. OpenAI's strategy relies on technological superiority rather than the market advantages like proprietary data and established customer relationships that its competitors possess.

— [Sam Altman Says, 'We Know How To Build AGI' — Not So Fast](https://www.forbes.com/sites/lutzfinger/2025/01/07/sam-altman---agi--ai-workforce-in-2025-the-battle-of-tech-giants/) · _Forbes_

Question 16

What does it mean for a company to be AI-native versus just using AI?

Accepted Answer

_AI-native means AI is the core product, not just an enhancement_

An AI-native company means the primary product offering or value is driven by AI, not just enhanced by it. Netflix is a channel for videos that uses AI to suggest movies, but you didn't need AI for Netflix. LinkedIn is a social network that uses AI to connect people, but you didn't need AI for LinkedIn. These are not AI-native. Companies like 11 Labs, Midjourney, or Cursor are AI-native because AI is at the heart of what's being built, what's being sold, and the people you hire early on.

— [The Future of AI, with Martin Casado](https://www.lutzfinger.com/transcript/2025-03-20-the-future-of-ai-with-martin-casado/) · _The Keynote on AI_

Question 17

What is Runway at Cornell Tech and how does it support PhD entrepreneurs?

Accepted Answer

_A two-year paid fellowship helping PhDs turn research into companies._

Runway is an entrepreneurship fellowship at Cornell Tech designed to help PhDs commercialize their research. The program accepts five fellows per year and provides a two-year paid fellowship specifically to commercialize technology that you ideally developed during graduate school. It's a structured way to transition academic research into viable commercial products, giving you financial support and time to build your startup while maintaining ties to Cornell Tech's ecosystem and resources.

— [Founder Coaching: Tips for your AI Journey](https://www.lutzfinger.com/transcript/2024-07-26-founder-coaching-tips-for-your-ai-journey/) · _The Edge_

Question 18

What is the minimum viable quality concept when building AI products?

Accepted Answer

_Building with AI means accepting imperfection strategically, not chasing perfection everywhere._

We always ask ourselves how to build a product where AI imperfection is not a problem for the use case you're solving. This is the concept of minimum viable quality, not product, but quality, because AI is not always perfect. Many startups put AI in the product but not actually in the solution. The value creation comes down to whether you have a decision that's different because you use AI. For example, a scale reminding you about your diet is just an interface advantage, not real AI value creation.

— [Building AI products: 5 lessons from our founders' workshop](https://www.lutzfinger.com/transcript/2024-08-01-building-ai-products-5-lessons-from-our-founders-workshop/) · _The Edge_

Question 19

What makes a data moat actually defensible in the LLM era?

Accepted Answer

_Ownership and unique value creation are both required for protection._

A data moat is created when the data you have creates value no one else can easily create, and you own the data. Google has a data moat because they know how to link a search query to a click, while ChatGPT does not know whether the user found an answer helpful. I failed with my own company Fisheye Analytics because our scraped data was not protected. If I can scrape it, anyone can scrape it. Bloomberg launched a massive financial LLM only they could have trained. In this new world, data access is key.

— [What Is The Competitive Advantage Of LLMs Like ChatGPT For Your Business? Three Takeaways.](https://www.forbes.com/sites/lutzfinger/2023/04/04/what-is-the-competitive-advantage-of-llms-like-chatgpt-for-your-business-three-takeaways/) · _Forbes_

Question 20

What makes companies investable in the current AI era according to Ben?

Accepted Answer

_Harder problems requiring deep industry insight are the new opportunities._

What used to be somewhat hard is not hard anymore, like pure workflow systems. The classic startup of a new SaaS replacing your finance process is gonna be tough. What I'm seeing more and more is people going after harder problems. Harder problems require deeper product insight and market knowledge. Ironically, in the world of AI, the clearer choices actually are worth it. All the bashing on SaaS was based on average workflow, but SaaS still has the database, privacy, enterprise scalability, customer support. You can plug any workflow on top, but that means you have to understand the industry deeply.

— [The Hidden Patterns, with Ben Orthlieb](https://www.lutzfinger.com/transcript/2026-04-23-the-hidden-patterns-with-ben-orthlieb/) · _The Keynote on AI_

Question 21

What makes Intuit's data moat unique compared to companies like Amazon, Google, and OpenAI?

Accepted Answer

_A unique combination of financial and customer data no frontier model possesses._

Intuit has data that Amazon, Google, and OpenAI do not have: 100 million customers whose tax filings, credit history, and business cash flow all live inside the same platform. Add Intuit Mailchimp and one gets a closed-loop financial plus marketing intelligence stack covering both the consumer and the small business. No frontier model has access to that combination. As Intuit's CEO Sasan Goodarzi put it directly in a February 2026 interview: data is the most important moat in all of this.

— [SaaSpocalypse Is Dead - The Future Of SaaS Is SaaS](https://www.forbes.com/sites/lutzfinger/2026/04/24/saaspocalypse-is-deadthe-future-of-saas-is-saas/) · _Forbes_

Question 22

What problems with RAGs still need to be solved before widespread adoption?

Accepted Answer

_The observability and control challenges that remain in RAG systems._

RAGs don't completely stop hallucination yet. The model uses both the trained corpus and new data, but we don't have a very good solution to manage that difference. Who decides when new information should override training data? Is the LLM taking that decision or the RAG? We lack observability and control. For example, I created a document saying I'm Captain America, but ChatGPT rejected it as fake. What if that was actually true breaking news? Amazing startups will help create this observability, solve access control, and manage the difference between what the corpus was trained on and corporate datasets.

— [Are we one step closer to using LLMs in practice?](https://www.lutzfinger.com/transcript/2023-07-21-are-we-one-step-closer-to-using-llms-in-practice/) · _The Edge_

Question 23

What specific regulatory changes does Europe need to make AI startups more viable?

Accepted Answer

_Legal shields and lower compliance burdens are critical for AI startups._

Europe needs several regulatory reforms. First, imagine a European equivalent of Section 230 for AI, a legal shield letting startups experiment without fear of lawsuits. Without it, regulation-heavy environments scare away founders. Second, lower compliance burdens. GDPR was a privacy milestone but became a speed bump for innovation. My own AI startup, r2decide, worked with a German e-commerce brand, but every advisor warned me to avoid launching in Europe due to compliance burdens. Even Apple delayed Siri upgrades in the EU due to regulatory friction. Europe must stop trying to retrofit old frameworks to this new paradigm.

— [Nvidia Wants To Power Europe's AI Dreams. The EU Isn't Ready](https://www.forbes.com/sites/lutzfinger/2025/06/16/nvidia-wants-to-power-europes-ai-dreams-the-eu-isnt-ready/) · _Forbes_

Question 24

Where does Ultimate AI see its competitive defensibility beyond just AI models?

Accepted Answer

_User experience and distribution trump data or model advantages_

I don't think there's a silver bullet or secret IP that guarantees you always win. On the data side, we stream live conversations from big companies in a secured way. We have hundreds of millions of proprietary conversations that competitors starting today don't have. But more important than data moats or technical moats is user experience and distribution. I love products that take 10 minutes to learn but years to master, like Shopify or Excel. If you build something easy to start but with depth, solid brand, and good distribution, that matters much more than some little advantage in the back office models.

— [How Ultimate evolved from hundreds of supervised models to UltimateGPT](https://www.lutzfinger.com/transcript/2023-05-12-how-ultimate-evolved-from-hundreds-of-supervised-models-to-ultimategpt/) · _The Edge_

Question 25

Which companies are competing in AI search right now?

Accepted Answer

_Major tech players and startups racing to dominate AI search._

Every major tech company and many startups are now racing to build or enhance AI-driven search. Google was late to the game, but its generative search experience is coming back quite strong. Microsoft has boosted Bing by integrating ChatGPT. Meta is developing its own search. OpenAI went from being a research organization to becoming a full-on consumer-focused search company. Perplexity focused on citation-based answers with source links. Many startups are also coming to AI-based search.

— [2025: The Year Of Search But Not The End Of Google](https://www.forbes.com/sites/lutzfinger/2025/01/03/2025-the-year-of-search-but-not-the-end-of-google/) · _Forbes_

Question 26

Who does Lutz expect will be Clubhouse's fiercest competitors and why?

Accepted Answer

_Twitter and Facebook have social graphs and content spread measurement._

I expect the fiercest competitors to be Twitter and Facebook, as they bring existing social network graphs and also measure how content spreads. As Clubhouse grows, their design choices will be necessary. Whether or not Clubhouse will become a super-aggregator will depend on their ability to create a multi-sided network with decreasing acquisition costs. In order to scale, they will need data and engagement metrics. The trend Clubhouse started is undoubtedly not a covid19 fad. It's the unbundling of content creation and distribution.

— [Clubhouse's Future Depends On Data - How To Build A TikTok Like Algorithm](https://www.forbes.com/sites/lutzfinger/2021/02/09/clubhouses-future-depends-on-datahow-to-build-a-tiktok-like-algorithm/) · _Forbes_

Question 27

Who is investing in these early-stage generative AI e-commerce companies you identified?

Accepted Answer

_Top VCs attend ShopTalk as startups drive generative AI innovation._

About 70% of the companies focused on data and AI at ShopTalk were startups, so it's no surprise that many top VC firms attend the event as well. I participated in a roundtable with Martin Casado from a16z at the conference. The fact that established VCs are present while most generative AI innovation comes from startups rather than established players, with only one out of the seven merged-solution vendors being an established company, shows where the smart money is watching this nascent space develop.

— [ShopTalk 2025: Just 1% Of Vendors Use Generative AI To The Max](https://www.forbes.com/sites/lutzfinger/2025/03/21/state-of-genai--only-1---shoptalk-2025-in-numbers/) · _Forbes_

Question 28

Why are radically lean teams the future?

Accepted Answer

_How AI compresses the team size needed to ship._

> But what hinders many of the large corporates to adapt them? >>Yeah, well, I think there's a lot of components there but in the end, it comes down to the size, and the politics, and the setup they built. Large companies are usually built to mitigate risk and focus on their core business. If there's a new technology that comes in that is not obviously threatening the core business, or opening a huge opportunity for the core business,…

— [Unit 5 Full Transcript](https://www.lutzfinger.com/ecornell/) · _eCornell course_

Question 29

Why are startups not yet widely adopting knowledge graphs with their LLM applications?

Accepted Answer

_Prompt engineering dominates now, but knowledge graphs will become essential at scale._

I am not seeing significant use of knowledge graphs at startups yet. There is heavy reliance on out of the box language models and significant investment in prompt engineering, both at startups and larger companies. The zeitgeist is to mine these wonderful resources rather than going upstream to drive the language models themselves. Knowledge graphs are not simple and are difficult to build, while prompt engineering and rule-based approaches are easier. Results from research literature take time to get absorbed by practitioners. Once startups scale or work on applications where accuracy is crucial, they will have to adopt knowledge graphs.

— [The synergy between LLMs and knowledge graphs](https://www.lutzfinger.com/transcript/2023-08-24-the-synergy-between-llms-and-knowledge-graphs/) · _The Edge_

Question 30

Why did you leave your healthcare company to start an AI e-commerce company?

Accepted Answer

_From healthcare prediction models to e-commerce AI startup founder_

I was president of a NASDAQ listed company in New York in the healthcare space, running product and technology for a third party administrator. We built predictive algorithms to reduce the burden of diseases for our members. I left to build a startup focused on fine-tuned large language models for e-commerce because I saw an opportunity. The technology behind it, transformer models, exists since 2017, but how everyone uses it creates new questions. I realized I could create value by knowing things the industry didn't know yet, specifically solving search and discovery problems in e-commerce with this new architecture better than anyone else.

— [After the Gold Rush, with Kurt Wagner](https://www.lutzfinger.com/transcript/2025-10-21-after-the-gold-rush-with-kurt-wagner/) · _The Keynote on AI_

Question 31

Why does Clubhouse's future success depend on using user data?

Accepted Answer

_Finding content on Clubhouse is hard without better data use._

Clubhouse's future success depends on how it can help users find content and enable engagement through the use of user data. At the core, any social app solves two issues: to find and enable social interactions. Currently, finding interesting conversations on Clubhouse is not easy. You curate your own content by following people, similar to how Twitter used the social graph as a first approximation for your interest. A social graph is a clunky way to curate your feed. In order to scale, Clubhouse will need data and engagement metrics.

— [Clubhouse's Future Depends On Data - How To Build A TikTok Like Algorithm](https://www.forbes.com/sites/lutzfinger/2021/02/09/clubhouses-future-depends-on-datahow-to-build-a-tiktok-like-algorithm/) · _Forbes_

Question 32

Why does Europe struggle with AI adoption compared to the United States?

Accepted Answer

_Underinvestment in training and punishing failure hold Europe back from AI._

Europe is not behind because it lacks brains. It is behind because it underinvests in training and adoption. In San Francisco, self-driving cars are a tourist attraction. In Europe, they're theoretical. Beyond training gaps, Europe doesn't lack risk-takers but penalizes them. In the U.S., failure is a badge of honor. In Europe, it's a career ender. We need policies like bankruptcy reform that give entrepreneurs a second chance. Europe has missed past digital revolutions including internet, cloud, mobile and social. Without ending the stigma of failure and accelerating adoption, infrastructure investment alone won't close the gap.

— [Nvidia Wants To Power Europe's AI Dreams. The EU Isn't Ready](https://www.forbes.com/sites/lutzfinger/2025/06/16/nvidia-wants-to-power-europes-ai-dreams-the-eu-isnt-ready/) · _Forbes_

Question 33

Why does Lutz agree with Macron that Europe is regulating itself out of the AI market?

Accepted Answer

_Europe risks losing the AI race through excessive regulation._

Macron said it correctly: "We are over-regulating and underinvesting. If we continue with our usual agenda over the next 2 to 3 years, we will be out of the market." I agree with him. Europe talks about regulating AI but misses the balance between fostering innovation and upholding societal values. Regulation doesn't happen in a vacuum. If we require the impossible or over-regulate while underinvesting, we risk stifling AI development entirely and losing competitiveness in the global market.

— [Emmanuel Macron On AI: We Are Regulating Ourselves Out Of The Market](https://www.forbes.com/sites/lutzfinger/2024/10/22/emmanuel-macron-is-right-about-ai-over-regulatingi-have-no-doubt/) · _Forbes_

Question 34

Why does Lutz think enterprises will adopt AI agents before consumers?

Accepted Answer

_Enterprises control risk and ROI better than consumer deployments._

I argue to forget fully autonomous AI assistants for consumers anytime soon. Enterprises will adopt AI agents first because they can control deployment, manage risks, and ensure ROI. Leading consulting firms like Accenture and BCG are already rolling out task-specific AI solutions. If we are there, Google and OpenAI will likely dominate this space, not Manus with same-old same-old LLMs. AI success is based on application, not "Made in China" branding. The future of AI isn't about flashy demos but about consistent, secure, and well-integrated solutions.

— [Overhyped: Manus From Monica — AI Made In China](https://www.forbes.com/sites/lutzfinger/2025/03/10/overhyped-manus-from-monicaai-made-in-china/) · _Forbes_

Question 35

Why doesn't OpenAI have a lasting model moat with ChatGPT?

Accepted Answer

_OpenAI's technical advantage disappeared faster than most expected._

OpenAI's model moat expired quickly after ChatGPT launched. We saw BARD, LLaMA, FLAN UL2, BLOOM, and then Stanford's Alpaca, a variant of Meta's LLaMA cross-trained by OpenAI's GPT-3.5. Everyone is now fine-tuning and cross-training models based on other models. A model moat does not exist for OpenAI. This is not surprising, such a moat also did not exist in previous Machine Learning waves. The only argument is that LLMs are expensive to train, so there's a barrier to entry, not a model moat.

— [What Is The Competitive Advantage Of LLMs Like ChatGPT For Your Business? Three Takeaways.](https://www.forbes.com/sites/lutzfinger/2023/04/04/what-is-the-competitive-advantage-of-llms-like-chatgpt-for-your-business-three-takeaways/) · _Forbes_

Question 36

Why have AI-native startups historically struggled to succeed, and what changed with generative AI?

Accepted Answer

_Why AI startups failed until generative AI changed the economics completely_

For 30 years, AI has been great technology but terrible business. The economics were never there. AI could improve existing businesses by 20 percent, which is amazing for Google or LinkedIn, but a 20 percent improvement doesn't make a startup. The big change with generative AI is it brings the marginal cost of creation and language reasoning to zero. For example, getting a Pixar-style avatar used to cost 200 dollars from a designer. Today, a model generates it for one one-thousandth of a penny. That's a five order magnitude cost difference, an economic dislocation that super cycles are made from.

— [The Future of AI, with Martin Casado](https://www.lutzfinger.com/transcript/2025-03-20-the-future-of-ai-with-martin-casado/) · _The Keynote on AI_

Question 37

Why is having a good DNA model not enough to build a defensible company?

Accepted Answer

_Models are commodities; the real advantage lies elsewhere_

The best DNA model available returns roughly 99% useless output. A model that can write DNA is not a company, it is a starting point. Selling a model does not create a moat. Training costs keep falling, open-source alternatives keep arriving, and Gartner now classifies foundation models as strategic commodities. The algorithm is necessary but almost never sufficient. When everyone has the model, the moat is customization through data and workflows, not the model itself.

— [DNA Is Becoming Programmable. Curing Cancer With AI.](https://www.forbes.com/sites/lutzfinger/2026/05/25/dna-is-becoming-programmable-curing-cancer-with-ai/) · _Forbes_

Question 38

Why is Microsoft considered to be in pole position in the race for AI agents?

Accepted Answer

_Why workforce tool dominance puts Microsoft ahead in AI agents_

Microsoft is in pole position because it has direct access to users through workforce tools most people already use. AI is becoming deeply integrated into these Microsoft products, with Copilots appearing everywhere. This market access is a big advantage. In 2023 and 2024, many startups launched impressive AI services, only to be quickly overshadowed by giants like Microsoft and Google, which have direct access to customers. For example, Jasper.ai saw similar features built directly into Google and Microsoft products.

— [The Race For AI Agents. Who Will Supply Tomorrow's Workforce?](https://www.forbes.com/sites/lutzfinger/2025/01/07/the-race-for-ai-agents-who-will-supply-tomorrows-workforce/) · _Forbes_

Question 39

Why is now a good time for founders despite potential market corrections?

Accepted Answer

_Excess infrastructure investment creates opportunities for practical applications._

As we come into those crunch times, you as a founder can get access to all this hardware and compute power for a much better future. We have too much supply of compute, and some people say there's 600 billion or 300 billion in excess hardware. People have invested and have to get their money back, so you can get better access to talent and resources. Even though I believe a correction will happen, the actual value we will see is way bigger. Plus, compute costs should drop massively in the future.

— [AI Gold Rush: Is the Market Overheating?](https://www.lutzfinger.com/transcript/2024-08-08-ai-gold-rush-is-the-market-overheating/) · _The Edge_

Question 40

Why weren't traditional banks serving the underserved population before Varo?

Accepted Answer

_Economic incentives and outdated systems excluded millions from banking._

There are two main reasons, focusing on economic incentives. It was really difficult for banks to make money off consumers with low-value, high-frequency transactions. Banks perceived them as higher credit risk because they used backward-looking systems instead of monitoring real-time cash flow. Vulnerable populations are also more susceptible to fraud and scams, creating high costs for dispute processing. The economic incentives made it very difficult for incumbent institutions to effectively serve this large population. Technology was needed to solve this.

— [Disrupting Traditional Banking, with Colin Walsh](https://www.lutzfinger.com/transcript/2025-05-22-disrupting-traditional-banking-with-colin-walsh/) · _The Keynote on AI_

Q&A, Startups

How can startups use large language models to improve their products?

How could Europe break down existing monopolies to help AI startups compete?

How do you decide if AI is the right fit for your product problem?

How does Lutz compare Clubhouse to the early internet and talk radio?

How does Neuralens measure brain health non-invasively compared to current methods?

How does your company Gooder AI address the gap between technical metrics and business value?

How does Zenith OS reduce costs in neuropsychiatry treatment?

How much capital was invested in the digital health startup landscape in 2021?

If OpenAI doesn't have a model moat, what kind of moat does it actually have?

Is ChatGPT actually at risk of shutting down despite the bankruptcy headlines?

Should founders focus on building new AI models or developing applications with existing technology?

What advantage do existing platforms like Google, Meta, and TikTok have in the bot space?

What advice do you have for someone considering building a fintech or launching in Africa?

What are the three key takeaways about competitive advantages with LLMs?

What competitive challenges will OpenAI face in 2025?

What does it mean for a company to be AI-native versus just using AI?

What is Runway at Cornell Tech and how does it support PhD entrepreneurs?

What is the minimum viable quality concept when building AI products?

What makes a data moat actually defensible in the LLM era?

What makes companies investable in the current AI era according to Ben?

What makes Intuit's data moat unique compared to companies like Amazon, Google, and OpenAI?

What problems with RAGs still need to be solved before widespread adoption?

What specific regulatory changes does Europe need to make AI startups more viable?

Where does Ultimate AI see its competitive defensibility beyond just AI models?

Which companies are competing in AI search right now?

Who does Lutz expect will be Clubhouse's fiercest competitors and why?

Who is investing in these early-stage generative AI e-commerce companies you identified?

Why are radically lean teams the future?

Why are startups not yet widely adopting knowledge graphs with their LLM applications?

Why did you leave your healthcare company to start an AI e-commerce company?

Why does Clubhouse's future success depend on using user data?

Why does Europe struggle with AI adoption compared to the United States?

Why does Lutz agree with Macron that Europe is regulating itself out of the AI market?

Why does Lutz think enterprises will adopt AI agents before consumers?

Why doesn't OpenAI have a lasting model moat with ChatGPT?

Why have AI-native startups historically struggled to succeed, and what changed with generative AI?

Why is having a good DNA model not enough to build a defensible company?

Why is Microsoft considered to be in pole position in the race for AI agents?

Why is now a good time for founders despite potential market corrections?

Why weren't traditional banks serving the underserved population before Varo?