The article provides an in-depth analysis of the views of top venture capital firms NFX and a16z, pointing out that in the AI era, speed has become a new moat for enterprise competition. The rapid development of AI has subverted the traditional entrepreneurial rhythm, and the speed has changed from a competitive advantage to a survival necessity. The article explores how AI is redefining speed and shares mindsets and strategies for moving quickly in the new era.
Have you ever thought that entrepreneurship may have changed completely?
I recently deeply analyzed the growth data of a large number of AI startups, studied the investor views of top investment institutions such as NFX and a16z, and discovered a shocking fact: we are at a historic turning point, where the traditional entrepreneurial rhythm, product development cycle, user growth model, and even the definition of “fast” have been completely subverted by AI.
When Perplexity AI went from zero to 15 million monthly active users and a $1 billion valuation in 18 months, when Cursor AI reached a valuation of $9 billion in less than two years with a team of 30 people, and when the monthly subscription fee for a consumer AI product jumped from $50 per year to $200 per month, I realized we needed to completely rethink what it means to be “fast enough.” What shocked me even more was that this speed is no longer an advantage for startups, but a basic threshold for entry.
If you’re still doing things with traditional entrepreneurial thinking and rhythm, you’re probably out of the game. For example, taking Cursor’s developer market as an example, if you want to make AI coding products for developers now, I think the opportunity is already slim.
Traditional enterprise-level startups that reach $1 million in annual revenue in their first year are considered excellent, while consumer companies often wait until they have millions of users before they start monetizing.
But now, these standards seem too conservative. According to a16z investors, the median annual revenue of AI enterprise companies reached over $2 million in the first year, while consumer companies performed even more astonishingly, with a median of $4.2 million.
What’s even more surprising is that the quality of the business model of consumer AI products has far surpassed that of traditional consumer products, with users willing to pay $20 per month for ChatGPT Plus and $250 per month for video generation tools like VEO, which completely subverts our perception of consumer product pricing.
The following is my recent research and analysis, I hope it will be helpful to you, just so it happens that I have published a separate article on the speed moat last week, “Speed will become the only moat in the AI era”, interested friends can read it.
How can product managers do a good job in B-end digitalization?
All walks of life have taken advantage of the ride-hail of digital transformation and achieved the rapid development of the industry. Since B-end products are products that provide services for enterprises, how should enterprises ride the digital ride?
View details >
Of course, everything is not absolute, non-consensus is precisely where the opportunity lies, as the saying goes, listening is clear, listening is dark, all my sharing is just a perspective, it is good to bring some inspiration to everyone.
AI is redefining the concept of speed
The most profound change I’ve observed is that AI is moving speed from a comparative advantage to an absolute necessity. In the past, speed could be a competitive advantage for some startups, but now it has become an essential competency that all companies that want to survive must have.
NFX investors pointed out in their analysis that the fundamental reason for this transformation is that AI itself does not hesitate or delay, and it can continuously ingest information, process data, and output results.
When your competitor may be an AI system or a team that uses a lot of AI, the traditional pace of work seems extremely slow and inefficient. As observed by a16z investors, we are living in the early days of AI, where speed is the moat.
NFX’s research shows that let me illustrate the scale of this change with some specific numbers. Mistral AI built Mistral 7B, a top-of-the-line open-source large language model, just four months after its inception. Replit integrated a complete toolchain of code completion, debugging, and deployment within 12 months.
Runway launched its second generation just eight months after releasing its first-generation video generation product, achieving a breakthrough in generating videos directly from text. And companies like 11Labs, according to the observations of a16z investors, went from the early days of the product to now with a huge voice library and a large number of enterprise customers, the entire process takes less than two years. These cases tell us that in the era of AI, the product iteration cycle has been shortened from years to months or even weeks.
NFX investors have found that the most successful teams don’t just build fast, but learn fast. They established a rapid iterative cycle between learning and release, which compounded in the dimensions of days, weeks, months, and years, forming a huge competitive advantage and potentially building a moat for the future.
This learning-release cycle is becoming a key metric that separates good companies from mediocre ones. When your competitors release new features, optimize user experience, and respond to market feedback every week, the traditional monthly or quarterly release cadence becomes extremely passive.
But I also realize that NFX investors point out a profound problem: most people don’t actually know that their speed thinking has been limited by past experiences.
What you learn in school is to advance by the semester, and what you learn is decision-making by committees at large companies, both environments teach you to try to avoid making mistakes. They reward not making mistakes, writing more words, deeper analysis, and being robust in a slow-moving system.
This logic works at the glacial speed of academic knowledge creation, but it doesn’t apply at all in the entrepreneurial world, especially in the age of AI.
NFX investors have observed that many entrepreneurs tie their self-worth to their products. They will think: I brought this product out into the world, and if people don’t like it, they don’t like me.
This unconscious mental connection makes them extremely cautious when launching products, overly pursuing perfection and procrastinating for fear of making mistakes.
They feel like they have a lot to lose, but in reality this thought makes them lose the most precious thing: time and opportunity.
Speed is not reckless, but a precise choice
I want to clarify an important misconception: speed does not equal recklessness. NFX investors observe that the truly successful startups are fast not because of chaos, but because of precision. They are fast because of clear thinking, they are fast because of clear goals, and they are fast because of smooth team communication.
Speed is essentially choosing motivation over perfection and progress over self-esteem. These companies are able to move fast because they have moved away from the excessive need for “security” and have carried the desire to move fast into every day of product development.
When I look back at all the entrepreneurial and company traits that NFX investors have summarized they are looking for, I see that they all end up converging on one metric: speed.
Whether it’s excellent communication skills, product obsession, resilience in the face of setbacks, deep market insight, or understanding of marketing, sales, and technology, all these elements that make a company successful are ultimately reflected in the dimension of speed.
When NFX investors see that founders are able to respond to email requests immediately, see them revise presentations quickly, see them launch products quickly, and see them interact quickly with customers, that speed is what they are looking for and one that everyone who wants to get into an entrepreneurial career should look at.
NFX investors see the tangible benefits of speed every day, not theory, but a phenomenon that actually happens.
The benefits of speed are mainly reflected in four aspects:
The product can be brought to market faster, the speed you bring features and the ability to change quickly for customers, customers can feel that rhythm, they will support you;
The team becomes more dynamic, the faster the team, the more excited the team is, and the more attractive you are to top talent, which in turn will make you faster;
You will get a lot of attention from the media and social media, the energy will gather towards you, and everyone can feel your rhythm;
You need to raise less money and keep more equity because you may spend $10-$300,000 a month, and if you can save three months, you’ll save $600,000 in expenses, and you’ll save yourself $600,000 in equity.
As NFX investors say, speed will compound, and hesitation will compound. You can choose your poison. Based on their observations in the current AI era, I have summarized six mindsets that must be shifted to enable entrepreneurs to move quickly in this new era.
A new business model for consumer products in the AI era
When delving into consumer-grade AI products, a16z investors have uncovered a striking phenomenon: consumers are now willing to pay unprecedented prices for AI products. Traditional consumer subscription products may cost as little as $50 per year on average, but now people are happy to pay $200 per month for AI products, and some users have even said they feel undercharged and are willing to pay more.
I think this change in price acceptance reflects a fundamentally different value offered by AI products: they are no longer tools to help you entertain or improve yourself, but assistants that can do the work directly for you.
As a16z investors point out, Deep Research, for example, can replace your own 10 hours of generating market reports, and for many, just one or two uses is worth the $200 per month fee. Video generation tools like VEO allow people to create content like never before, allowing users to craft personalized video messages and create complete stories, which justifies the high price.
I’ve observed that the value proposition of AI products has shifted from “help you do something better” to “do something directly for you,” a fundamental shift that explains why consumers are willing to pay more.
What’s even more interesting is that a16z investors found that consumer-grade AI products show a different “shape” from traditional consumer products. According to their research, one-third of consumer companies in the sample raised significant funds to train their own models, and many see a huge jump in revenue after the new model is released.
These increases often take the form of a step function, potentially leading to a plateau before the next major launch and then jumping again. This model is completely different from the steady growth of traditional consumer products, and is more like the release cycle of technology products.
a16z investors also noticed an important phenomenon: in consumer-grade AI products, they began to clearly distinguish between user retention and revenue retention, which was rare in previous consumer products. Because users actually upgrade their plans, purchase more credits, and incur overage fees, revenue retention is often significantly higher than user retention. This means that even if some users drop off, the remaining users generate more value, which is a very healthy business model signal.
From an investment perspective, a16z investors have observed that the quality of the business models of these consumer-grade AI products far exceeds that of traditional consumer products. The top tier plan for ChatGPT is $200 per month, and the top tier plan for Google’s consumer products is $250 per month, a level of pricing that is unthinkable in past consumer products.
What’s more, these companies are raising prices rather than reducing them, indicating a strong demand for high-quality AI products and that supply remains scarce.
Socializing and connecting: new opportunities in the age of AI
In my research with a16z investors, I found an interesting point: in the current wave of AI products, we have seen a lot of innovation in information processing (such as ChatGPT), utility tools, and creative expression tools, but there is a relative blank space in terms of connection and socialization.
Traditional social networks such as Facebook, Instagram, and Twitter are almost all products from 20 years ago, and new AI-based social experiences have not really emerged. a16z investors believe this may be because most AI products are developed by research teams that are good at training models, and they have relatively limited experience at the consumer product level.
One phenomenon that I find particularly interesting is that a16z investors have observed that people are starting to “confide” more personal information to ChatGPT than to Google.
Many people may have used Google for more than a decade, but ChatGPT may have outperformed Google in understanding them because people actively input more, provide more context, and share more personal ideas.
This deep collection of personal information opens up new possibilities for future social products: what kind of connection experience would this “personal essence” be shared with others?
a16z investors have noticed that we have seen some early indications, such as people asking ChatGPT to write five pros and cons based on what they know about them, or create an image that represents their essence, or make cartoons about their lives and then share them on various platforms.
This self-expression based on AI understanding is becoming a new social behavior, but it is still mainly happening on existing social platforms rather than on new AI-native platforms.
Voice interaction could be an important breakthrough in social products in the AI era. a16z investors pointed out that voice is the basic medium of human interaction, but only now has the technology truly matured to support natural voice interaction.
They see a lot of products starting to explore voice modes, like ChatGPT’s advanced voice capabilities, and people are using it for all sorts of things
Six mindsets that must be changed
Research based on hundreds of AI companies summarizes six mindsets that must shift in the age of AI.
The first mindset shift is: the product is not you. NFX investors emphasize the need to stop tying self-worth to products. Launch a product, see how people use it, and move on to the next thing. Your position in seven to ten years is the measure, that’s how you measure your life.
I’ve seen too many entrepreneurs hesitate to release their products because they fear they won’t be accepted, but in fact, early user feedback is the most valuable resource that can help you quickly adjust the direction and find the real product-market fit.
The second thinking is: you are not running a business, you are conducting a series of experiments. NFX investors note that if you have this mindset, things will go a lot faster. Treat everything you do as a simple test, and you’ll learn more and don’t care as much about whether it’s right or not.
This experimental mindset allows you to try and learn quickly instead of spending a lot of time on perfect planning. Facebook, Google, Airbnb, none of the companies you admire were built from scratch, they all borrowed something and components from other effective companies and did it better.
The third thinking is: copy what works. NFX investors emphasize that you don’t get extra points for building from scratch. Your brain is different, your team is different, you are in different moments, and you use different technologies.
Don’t worry, just because you are you, in the times you live in, you will end up getting something unique. Don’t be original for the sake of originality, learn for the sake of effect. The smartest entrepreneurs know how to stand on the shoulders of giants and then add their own unique value.
The fourth thinking is: let go earlier. NFX investors sometimes call it “fucking moment”. By holding on too tightly, you often don’t allow the best ideas to come out. When you let go, when you don’t hold on as tightly, when you don’t always try to get it right, the best ideas come and you end up doing the right thing. So let go earlier. Perfection is the enemy of good, and excessive control can hinder innovation.
The fifth thinking is: avoid self-sabotage. NFX investors have observed that entrepreneurs are often driven by some kind of fanatical reason to be obsessed with building something in the world, creating something.
But behind many of these drives, there is some tendency to self-sabotage. Be aware of these ways and avoid them. Many times, our worst enemies are our own fears and limiting beliefs.
The sixth thought is: realize that you are doing it for someone else – your employees, your family, your customers. NFX investors point out that if for some psychological reason you can’t get speed for yourself, or if there’s something fear standing in the way, you need to realize that you have a higher mission and you need to do it for them. When you shift your focus from yourself to serving others, many psychological barriers will naturally disappear.
How to get started right away
Based on my observations and advice from NFX investors, there are five specific actions that can immediately bring speed thinking to you and your team. First, redefine what “fast” means with your team. Directly announce that you will adopt different ways of acting and different rhythms, and encourage team members to supervise each other and push each other faster and faster.
This is not a one-man battle, but a cultural shift for the entire team. In the success stories I research, teams that can respond quickly to market changes tend to share this common culture of speed.
Second, NFX investors suggest releasing that 80% finished stuff today. Get feedback, maybe publish under a different name, at a different URL, but do something to get it out.
I’ve seen too many entrepreneurs wait for the so-called “perfect timing,” but in reality, 80% of solutions are released today, which is much better than 100% of solutions released next month. Market feedback will tell you which direction the remaining 20% should go. Judging from the data of a16z investors, those companies that can iterate quickly often achieve revenue several times higher than the industry average in the first year.
Third, NFX investors emphasize not to over-plan. Release, learn, and iterate quickly. Traditional businesses may require detailed planning and lengthy decision-making processes, but in the age of AI, the market changes too quickly, and overplanning often means missed opportunities.
A better strategy is to release a basic version quickly and then iterate quickly based on user feedback. The successful companies I’ve observed, from Mistral to Cursor, have adopted this rapid iteration approach.
Fourth, NFX investors suggest that AI be your co-founder. Think of AI as your colleague. Let it remove your bottlenecks and become part of the team. AI is more than just a tool, it should be seen as a team member who can work 24/7, without fatigue, without getting emotional.
It can handle a lot of repetitive work, freeing up human team members to focus on more creative and strategic tasks. In my research, companies that deeply integrate AI into their workflows develop products 3-5 times faster on average.
Fifth, NFX investors suggest having someone on your team dedicated to implementing new AI-based systems that work on your behalf. Let them spend their weekends and evenings exploring possibilities and implementing a new system every week, every two weeks.
After 25 weeks, you’ll have a lot of failed systems, but you’ll also find amazing systems that save you a lot of time and allow you to serve your customers faster. This continuous technological exploration and experimentation is key to maintaining a competitive edge.
Remember, NFX investors emphasize that you don’t always have to feel ready, you don’t always have to be right, but you have to stay active. As investors, this is what they look for in founders – those who are in motion. When the window of AI opens for startups, the cost of hesitation is higher than the cost of failure, because AI has no fear, AI does not hesitate, and this is the new standard.
New hardware forms and future opportunities
When discussing the future of the AI era, a16z investors believe that innovation in hardware forms could lead to the next wave of significant opportunities. With 7 billion mobile phones worldwide, few devices reach this level of popularity. But I observed some interesting trends: a16z investors noticed that young people were starting to wear devices capable of recording conversations at tech meetups, and they found real value in this. This concept of an “always-on” AI companion is taking place from sci-fi to reality.
The a16z investor pointed out that AirPods may be the most widely adopted device after mobile phones, which makes me feel that it hides great potential. While there are currently some social etiquette issues (like wearing AirPods to dinner is considered impolite), new social norms may evolve to accommodate AI’s ongoing interactions with us. Imagine your AI assistant being able to hear all your conversations, see everything you do, and then tell you, “If you spend an extra 5 hours a week doing this, you can become a world expert in this field,” or “Based on the vast network of users I serve, you should connect with these three people, and this person could be your perfect co-founder.”
a16z investors are particularly interested in products that can see what you see on your screen and act on your behalf. These products not only provide advice, but also actually perform tasks, such as sending emails, scheduling meetings, etc.
As the agent model becomes more powerful, this shift from suggestion to actual execution will bring immense value. Instead of passively responding to your requests, future AI assistants will actively observe your behavior patterns, anticipate your needs, and start preparing solutions before you even realize it.
Voice interaction plays a key role in this new form of hardware. a16z investors have observed that voice is the foundation of human interaction, but it is only now that the technology has truly matured.
They are seeing businesses adopt voice AI to replace human agents on a large scale, not only in low-stakes conversations, but even in sensitive areas like financial services. This indicates that the quality of voice AI has reached a level that can handle critical business conversations.
These hardware and interaction innovations will provide new opportunities for entrepreneurs. Those teams that can seize these new forms and new ways of interaction are likely to become the next generation of platform-level companies. But speed is still key – you need to build and capture the market quickly before these new opportunities are fully recognized and competitive.
The window period closes faster than you think
I must emphasize a stark reality: NFX investors point out that this window closes faster than you think, always. This is not just a motivational speech, it is reality. If you want to create something valuable and interesting in the age of AI, you must accept this reality and act accordingly. The data I’ve seen shows that companies that caught the AI wave early on are widening the gap with latecomers at an unprecedented rate.
From the hundreds of company data analyzed by a16z investors, we are indeed in a new era of entrepreneurial growth. The median in the enterprise-level company sample achieved annual revenue of more than $2 million in its first year and completed its Series A funding round just nine months after it began profitability.
Consumer-grade companies performed even better, reaching $4.2 million in annual revenue and completing a Series A funding round in eight months. The growth rate of annual revenue from zero to $1 million, once considered “top-notch,” is now at the lower end of the growth we’re seeing.
What impressed me even more was that a16z investors observed that the gap between top performers and average performers was widening. Many breakthrough companies continue to accelerate in the first year rather than starting to slow down growth as we often saw in the pre-AI era. This shows that there is a huge demand for excellent products, whether it is enterprise users or consumers, so it is worth trying boldly.
Of particular note is that a16z investors have found that consumer companies are now truly profitable. Somewhat surprisingly, B2C revenue benchmarks are outpacing B2B. This is partly because consumer companies now have a different “shape”. One-third of the consumer companies in their sample raised significant amounts of money to train their own models, and many saw huge revenue growth after the new model was released. These surges often resemble step function growth, potentially leading to a plateau before the next release.
While generative AI B2C businesses may have lower paid conversion rates than their pre-AI counterparts, data from a16z investors shows that once users do convert, their retention rates are just as good. This means that while acquiring paying users may require different strategies, once acquired, user value is long-lasting.
Another interesting phenomenon: B-end companies tend to be slower to adopt new technologies, but this is changing. The Product Hunt founder predicts that next year we will see this reversed, with enterprise-grade AI applications surpassing consumer-grade applications in terms of revenue growth. This presents a huge opportunity for entrepreneurs who are now focused on the enterprise market, but only if they can move quickly and establish a dominant position before the market is fully mature. This point should be paid special attention to companies going overseas, and overseas is still the world of the B-end enterprise market.
In this new era, if you’re still using traditional entrepreneurial thinking and rhythm, if you’re still waiting for the “perfect timing” or “perfect product,” you may have missed out on the most important window of opportunity. Entrepreneurship in the age of AI is not about waiting, it’s about acting; Not about perfection, but about speed; It’s not about avoiding mistakes, it’s about learning and iterating quickly. Remember, as NFX investors emphasize, in this day and age, speed is not just an advantage—it’s a fundamental condition for survival.
What does a product manager need to do?
In the process of a product from scratch, it is not easy to do a good job in the role of product manager, in addition to the well-known writing requirements, writing requirements, writing requirements, there are many things to do. The product manager is not what you think, but will only ask you for trouble, make a request:
View details >
A new model for enterprises to adopt AI: penetration from the consumer side
a16z investors observed a surprising phenomenon: companies sometimes adopt certain AI products earlier than consumers, which is a stark departure from previous technology proliferation models. In the case of 11Labs, a16z investors found that the company initially attracted early adopters and consumers who used it to make memes, funny video audio, and game mods. But then, this company has secured a large number of enterprise contracts and huge enterprise customers covering multiple use cases such as conversational AI, entertainment, and more.
What’s interesting about this pattern is that enterprise buyers now have a strong need for directives and AI strategies to use AI tools, and they actively follow Twitter, Reddit, and various AI newsletters to find seemingly random consumer-grade meme products and then think about how to apply them in their own businesses. This kind of active learning behavior of enterprise decision-makers is rare in previous technological waves.
What’s even more surprising is that a16z investors have also discovered a new way to acquire enterprise customers: by analyzing Stripe payment data to find out if a company has a large number of employees using a certain AI product, and then reaching out to the company to ask if they need enterprise-grade services.
In the music and creative sectors, a16z investors believe we are facing a critical problem: AI-generated content often appears “mediocre.” This is because AI is essentially an averaging machine, and culture should be on the margins, embodying uniqueness. The real question is not AI creators vs human creators, but bad art vs good art.
If AI can create art of the same quality, people may not care if the creator is human. But the point is that if you train a model with all the music that existed until hip-hop, it may not be able to infer hip-hop because music is the intersection of past music and culture, and culture is essential to creating new and interesting music.
a16z investors have also noted the evolution of the concept of moat and defensibility in the AI era. Traditionally, network effects, being part of a workflow, being a system of record, etc. have been important moats. While these are still important, they found that companies and investments that were examined with the moat-first theory did not actually perform as expected.
The real winners are often those that break the mold, move fast, and have amazing model launches and product iterations at speed. In the early days of AI, speed was a moat, both in terms of distribution (it’s hard to cut through the noise) and in terms of product velocity, which led to a share of thought that translated into users and traffic, and ultimately real revenue.
AI’s reshaping of the creator economy and the entertainment industry
a16z investors have deep insights into the future of the creator economy. They believe that we will see a divergence between creators and celebrities: one is people like Taylor Swift, whose human experience matters, where people not only love her songs, but also resonate with what happens in her life, her story, her live performances, etc., which AI cannot replicate at the moment. The other category is interest-based creators or celebrities, like the ChatGPT conversation we talked about earlier with Thomas Train, where it doesn’t matter if the person has real human experience, what matters is whether they can have an interesting conversation or share content around a particular topic.
When it comes to video content creation, a16z investors have observed some exciting trends. They noticed that tools like VEO3 unlocked new creative possibilities, such as street interview format videos, but the interviewees could be elves, wizards, ghosts, or furry characters that younger generations like.
This content creation of non-human characters opens up entirely new forms of entertainment. We’ve seen an explosion of AI influencers, but there are still very few who can actually become top influencers like Little Meloya, suggesting that creating great AI art still requires a lot of time and skill, just like traditional art creation.
a16z investors particularly emphasize an important point: while it is now easier for anyone to generate art than before, it still takes a lot of time to produce truly good AI art. At their AI artist events, many artists demonstrated their workflow for making AI films, which may actually take comparable time to traditional shooting, but these artists may not have previous shooting skills, so AI allows them to achieve things that they couldn’t do before. This democratization effect will generate a lot of AI talent and human talent, and the best will stand out, and the conversion rate may be low, but that’s exactly what it should be.
The moat of the AI era redefined
Discussing the defensibility of AI companies, a16z investors offer a thought-provoking perspective. They note that while all base models may look interchangeable to some extent, whether that means price pressure, different people use them for different purposes, and that these models are actually raising prices rather than decreasing them. This suggests that even in seemingly commoditized fields, there is still room for differentiation and value creation.
In the case of 11Labs, a16z investors observed an interesting network effect. When they made AI-generated videos that required dubbing, 11Labs had better models because of the lead, and more people used the product, making the models even better, while accumulating a large library of user-uploaded voices and characters.
So, when a very specific voice is needed, like an old mystical wizard voice, 11Labs may have 25 options, while other platforms may only have 2-3. This data network effect is similar to that of traditional markets but is built on AI capabilities.
a16z investors have also observed an important phenomenon: many successful AI companies are experiencing a “snapshot gingerbread man strategy.” The concept comes from a blog post from a decade ago, and the basic idea is that “anything Snap can do, Facebook can do better, but Snap will continue to come up with the next innovation.” The same is true in AI, where the key lies in continuous innovation and rapid iteration, rather than trying to build a one-time moat.
But I think the most important observation is that a16z investors emphasize that distribution and network effects will eventually come into play. Snap also has networking in its own field, with a younger user base as a corner of its core messaging platform. In AI products, we have not yet seen the emergence of real network effects, mainly because most of them are currently creative products, and there is no closed-loop social network that has not formed a creation-consumption-network effect. But once this closed loop is formed, it will create an extremely strong competitive advantage.
The redefinition of “work” in the age of AI
As I delved into the data of a16z investors, I realized that we needed to redefine what “work” meant. In the traditional software era, “work” means that users need to learn the interface, remember the operation steps, and adapt to the software logic. But in the age of AI, the real “work” is directly producing results and value. a16z investors have found that the most successful AI applications are not those with the most functions, but those that can complete tasks for users most directly, which is also my view – products must be closed-loop.
I observed three notable changes: from click to conversation, from learning to expression, and from process to outcome. Users no longer need to click a dozen buttons to complete a report, they just need to say “help me analyze last quarter’s sales trends.” Users no longer need to learn complex software operations, they just need to clearly express their needs. Users no longer care about how the software works, they only care about getting the desired results.
Data from a16z investors shows that this shift is creating unprecedented business value. Companies that understand this shift are growing at a rate that traditional software companies can’t imagine. The median revenue for enterprise AI companies reached $2 million in the first year, and for consumer companies reached $4.2 million, a figure that would take 3-5 years to achieve in the traditional software era.
What’s more, a16z investors found that the retention and expansion rates of AI applications far exceeded those of traditional software. Because when software is truly able to “work” for users rather than “work” them, users naturally rely more on and trust these tools. This explains why consumers are willing to pay 4-10 times more for AI products than traditional software.
What this means for founders
Based on my in-depth analysis of these trends, I believe that the current entrepreneurial environment places three fundamental requirements on founders. The first is a faster pace of financing. Data from a16z investors shows that the median time for AI companies to complete Series A financing is 8-9 months, more than double that of traditional companies. This means founders need to start preparing for financing earlier and prove product-market fit faster.
The second is a new understanding of the gap between “excellent” and “excellent”. a16z investors have observed that while the overall bar is improving, the top performers are leading significantly. Many breakthrough companies are not only not slowing down but accelerating their growth in the first year. This shows that there is a huge demand for excellent products in the market, and it is worth pursuing higher goals for founders. At the same time, it also means that if you are not making a truly exceptional product, it will be extremely difficult to compete.
The third is the fundamental change in the business model. Consumer-grade AI companies are now truly profitable, rather than having to accumulate a large number of users before monetizing like traditional consumer products. This opens up new opportunities for founders, but also requires them to think about how to create real value from the start, not just capture user attention.
I would especially like to emphasize that speed has gone from a competitive advantage to a barrier to entry. NFX investors point out that teams that are still working at a traditional pace, even if the product is good, may miss out on market opportunities because they are too slow. In an era where AI is tireless and doesn’t hesitate, human teams must match or even surpass AI speed to win the competition.
From a practical perspective, I would recommend that founders revisit their product development processes. The traditional “plan-develop-test-release” model is too slow and requires a shift to a fast “build-measure-learn” cycle. At the same time, use AI as a team member, let it take on a lot of repetitive work, and unleash the creativity and judgment of the human team.
Ultimately, I think successful founders will be those who can achieve high speed while maintaining high quality. This is not an either-or-one problem, but a challenge that must be addressed at the same time. As NFX investors say, speed will compound, and hesitation will compound. The choice is in the hands of the founders.
In this new era, if you’re still using traditional entrepreneurial thinking and rhythm, if you’re still waiting for the “perfect timing” or “perfect product,” you may have missed out on the most important window of opportunity. Entrepreneurship in the age of AI is not about waiting, it’s about acting; Not about perfection, but about speed; It’s not about avoiding mistakes, it’s about learning and iterating quickly. Remember, as NFX investors emphasize, in this day and age, speed is not just an advantage—it’s a fundamental condition for survival.
My deep thoughts on the nature of entrepreneurship in the age of AI
After delving into the perspectives of these investors and numerous cases, I began to think about a deeper question: what we are experiencing is not just a technological upgrade, but a fundamental restructuring of business civilization. The traditional business world is built on scarcity – time is scarce, talent is scarce, resources are scarce, and AI is systematically eliminating these scarcities. When tasks that once required a lot of manpower and time, such as content creation, code writing, and data analysis, can be completed by AI in an instant, the logic of value creation is bound to change fundamentally.
I realized that in this new era, what is truly scarce is no longer the ability to execute, but judgment and a sense of direction. Anyone can get AI to generate 10,000 ideas, but those who can identify which ideas are truly valuable will have a huge advantage. Anyone can get AI to write perfect code, but the people who can design the right product architecture will dominate the market. This shift from execution-oriented to insight-oriented requires us to rethink the logic of education, training, and even career development as a whole.
On a deeper level, I think we are entering an era of “intent economy”. In the past, our economy was built on the exchange of goods and services, with people buying specific products or services. But in the age of AI, people are increasingly buying results and experiences. Users don’t care how many parameters your AI model has, they don’t care how complex your algorithm is, they only care about whether you can help them achieve the results they want. This shift from process-oriented to result-oriented will completely reshape business models across all industries.
I’m also thinking about a more philosophical question: Where is the unique value of humanity when AI can do more and more “intellectual work”? I conclude that human values will increasingly focus on three dimensions: creative imagination (proposing unprecedented possibilities), emotional connection (building genuine relationships), and moral judgment (making the right choices in complex situations). These three dimensions are precisely the most difficult for AI to replicate at present, and they are also the capabilities that future education and personal development should focus on.
Ultimately, I believe we are in the most exciting moment of innovation in human history. Just as the Renaissance unleashed human artistic creativity, the Industrial Revolution unleashed human productivity, and the AI revolution is unleashing human intellectual creativity. Those who can understand the nature of this release and redesign their way of thinking, working, and even living accordingly will have unprecedented opportunities in this new era. And those who are still thinking about the problems of the new era with the logic of the old era are destined to be left behind by the wheels of history. Speed is not just a requirement for success, but a necessity for evolution.
