This shift means a reassessment for both large companies and startups. Apple's risk lies in the potential weakening of its long-standing software experience premium if the interaction layer is taken over by AI agents; for SaaS companies, the risk is that functionality itself is becoming increasingly difficult to use as a competitive advantage.
However, the democratization of software production capabilities may also lead to a new wave of individual creators and small team companies. This is a dangerous era for homogenized software; for founders with distribution, taste, resources, and industry depth, it may also be an unprecedented window of opportunity.
Apple is dead; the market just hasn't had time to complete the formalities for it yet.
This isn't just a sensationalist judgment, but a structural summary of the industry changes over the past six months. Navarre Lavikant's remarks on his podcast last week almost confirm this. One of the most patient investors in the tech world, and one of the most astute capital allocators of the past two decades, has given an extremely clear conclusion about the entire software industry: pure software is no longer worth investing in.
For founders, the real question is not whether they agree with the assessment, but whether they have 18 months to complete the transformation before the market fully reacts.
Background: Naval founded AngelList and was an early investor in Twitter, Uber, Notion, and about 200 other companies that shaped the tech landscape of the past decade. He rarely makes judgments lightly, but when he does, they are often cited repeatedly for many years. Therefore, when he says "pure software isn't worth investing in," it's not a casual comment, but a capital allocator's repricing of industry cycles.
Here is his assessment, and what this means for all entrepreneurs.
Apple won't go bankrupt, and it won't disappear from your pocket next year. The collapse Naval is referring to isn't operational, but economic.
Apple's $3 trillion market capitalization is fundamentally based on only one thing: using superior software experiences to support the premium pricing of high-end hardware. Once this experiential advantage ceases to hold, Apple will become a more refined Samsung. And this is already happening.
Interaction layers are being commoditized. In the next 24 months, the way most people open apps will change: they will no longer actively enter individual apps, but will directly interact with AI agents, which will instantly generate the required interfaces. Apple's carefully constructed App Store, human-computer interaction standards, design aesthetics, and ecosystem moat will quickly lose their original value once the interfaces themselves can be instantly generated by AI on any device.
What was Apple's response to this change? They licensed the technology to Google and introduced Gemini.
This means that the company that has always considered "controlling the user experience layer" as its core identity is outsourcing that layer to its strongest competitor. After its bet on developing its own AI failed, Apple is using external models to fill the gaps in its internal strategy.
This is almost a rapid replay of the "post-mobile Microsoft" script.
Microsoft missed the mobile era not because it lacked resources, but because it was unwilling to build a native touch-based operating system from scratch. Its dominance in the old era led it to mistakenly believe that the old norms would continue. By the time Microsoft truly faced reality, Apple had already won the next decade. Today, Microsoft is still a $3 trillion company, but Windows has lost the consumer war it could have won.
Apple is now making the same mistake in the AI wave: it still believes that its hardware-first DNA can carry it through the era of intelligent agents.
But this path is destined to be difficult. Once the operating system and user interface are commoditized, Apple's profit margins will be compressed to the level of hardware products. And hardware premiums are precisely the core source of profit supporting Apple's entire business empire. At that point, a structural revaluation of revenue and valuation will be difficult to avoid.
You can certainly continue to hold Apple stock, but stop treating it as a growth stock. This most valuable hardware company in history is about to be forced to answer a brutal question: without its software moat, how much is its hardware really worth?
For the founders, the more difficult part to accept is this.
Navarre's statement that "pure software isn't worth investing in" is correct in itself. However, what he didn't elaborate on is: what will happen to those SaaS companies that raised funds at Series A or Series B valuations in the previous funding cycle?
The answer is: most of them are already dead, they just don't realize it yet.
The logic isn't complicated. Your SaaS company exists because building this product was difficult in the past. You were able to raise funds because executing the technology requires a complete team. Your moat—whether you like it or not—essentially comes from the difficulty of replicating what you've built.
And this difficulty is collapsing.
A two-person team using Claude Code can replicate 80% of the core functionality of most B2B SaaS products within 90 days. This isn't a toy version, but a usable product with a reasonable architecture, basic security, and room for expansion. The remaining 20%—specific integrations, enterprise sales systems, and compliance processes—still exist, of course. But that's not a moat; it's more like friction costs. And with the next generation of intelligent agents iterating every quarter, these friction costs will continue to be reduced.
Similar changes are already emerging. Adobe acquired Figma for $20 billion in 2022 because Figma was considered a structurally difficult-to-replicate product at the time. But now, design tools with 70% of Figma's core functionality have been created by independent developers within months.
Salesforce is one of the most valuable SaaS companies in history. But AI-native CRMs, which didn't exist just 18 months ago, are already eroding its market share in the mid-range market. Workday, ServiceNow, Atlassian, Asana—each is becoming a potential target for AI-native alternatives, and the teams behind these alternatives are even smaller than their own HR departments.
The companies that will survive this transformation will not be those that write the best software. Because the value of software itself is approaching zero.
Those companies that will truly survive are those that have built something that AI cannot directly replicate: distribution channels, network effects, data flywheels, hardware integration, branding, community, and regulatory barriers. These are the only remaining lasting lines of defense in the new era.
If your honest answer to the question "What is our moat?" is "Our product is better," then you probably only have 18 months to find a real moat. Otherwise, you may very well watch your valuation evaporate by 70% to 90% in the next round of funding.
Founders who survive this transformation are those who take these signals seriously today. Those who choose to ignore them are highly likely to write a layoff letter in 2027, then ask in bewilderment: Why did everything happen so fast?
The question is, which one are you?
If pure software is no longer worth investing in, then what is worth investing in?
In his podcast, Navarre offered guidance: hardware, AI models, and businesses with network effects. Expanding further, what founders truly need to consider right now are the following types of competitive moats.
First, pipeline distribution.
Today's truly dominant companies are not necessarily those with the best products, but rather those with the most direct relationships with their customers. Products are merely the vehicle for serving customers; the audience is the true competitive advantage. Your email lists, communities, reputation, and distribution networks are all assets.
If you still believe that "marketing" is a stage that only begins after the product is perfected, then you are already behind the times. In the future, marketing itself will be part of the product; the product will merely be a downstream channel for generating traffic and relationships.
Second, the network effect.
Businesses that can resist the commodification of AI are those whose value comes from the users themselves, rather than the features themselves. Discord, Roblox, LinkedIn, and Reddit cannot be easily replicated, not because their software engineering is so complex, but because their users are locked into them by other users.
Does your product become more valuable as your user base grows? If the answer is yes, you have sustainability. If there's no fundamental difference in product value between 100 users and 100,000 users, then you're in danger. AI can replicate functionality, but it cannot replicate a truly functioning community.
Third, the data flywheel.
Only companies that can accumulate proprietary data through user interaction and use this data to train better models and form a feedback loop possess long-term value. Tesla's Autopilot data and Bloomberg Terminal data are essentially growing at a compound interest rate.
However, if your product is merely a UI layer wrapped around a public API, then you have no real assets. If each user interaction doesn't generate data that competitors can't access, your product will struggle to establish a long-term competitive advantage.
Fourth, hardware integration.
Companies that control the physical layer have the longest defense cycle. Tesla, Anduril, SpaceX, Apple's chip business, and Boston Dynamics are all typical examples. Hardware is difficult, the supply chain is difficult, manufacturing is difficult, and the complexities of the physical world cannot be directly smoothed out by AI.
AI will not automatically manufacture chips, batteries, rockets, or robots. The physical world remains one of the most difficult moats in the entire economy to replicate quickly.
Fifth, vertical depth.
Horizontal SaaS giants have the greatest risk exposure, while vertical platforms that are truly deeply rooted in the industry are actually safer. General project management tools are already very risky, but if you are deeply rooted in the construction industry, controlling approval processes, checking networks, regulatory data, and industry relationships, it's a different story.
In the future, it's better to go deep enough in one industry than to make superficial tools in ten industries.
If you are restructuring your strategy right now, there is only one core question: What kind of real moat can you build in your business over the next 12 months? Not some future day, but right now.
The founders who are the first to complete the transformation will take over the survivor market after others have fallen.
This is also the part that many founders most easily overlook when they hear "software is dead." They only see what is being destroyed, but they don't see the opportunities that are being created.
Navarre's most optimistic assessment on the podcast is that software is experiencing a renaissance for individual creators. This is not the death of software, but rather the democratization of software production capabilities.
Similar histories are not without precedent. Notch developed Minecraft single-handedly; Markus Frind built Plenty of Fish into a company with annual profits of $10 million on his own; when Instagram was acquired by Facebook for $1 billion, the entire company had only 13 employees; and when WhatsApp exited the market for $19 billion, it had only 55 employees.
These companies have collectively demonstrated one thing: a founder's vision that is not diluted by organizational coordination costs can directly lead to product implementation.
In the past, however, they were more often seen as outliers. Independent founders could create interesting things, but struggled to overcome the hurdles of scaling. Once the company expanded and the team swelled, compromises began to emerge, and the vision started to be diluted. That unique element that made the product stand out often slowly disappeared during committee-style refinement.
What's really changing now is the ceiling.
Navarre envisions a future where a one-person company can operate at the speed of a 50-person team. Users report bugs within the app, and the AI automatically reviews them every 24 hours, writes fixes, submits pull requests, and executes tests; the founder only needs to audit, approve, and deploy. Customer support is handled by the AI, which can also reverse engineer code to fix underlying issues. Users vote on feature requests, the AI is responsible for building them, and the founder is responsible for quality control.
There were no coordination costs, no internal politics, no diluted vision, no engineers passing the buck on key details, no designers arguing over the placement of graphics, and no product managers turning a bold version into a safe one.
The founder's vision can go directly from the brain to the top, with almost no organizational losses in between.
This isn't theory; it's already happening regionally. Pieter Levels, operating independently, has built multiple seven-figure revenue businesses. A growing number of independent developers are running companies that, three years ago, required Series A funding to survive. AI-native independent operators are creating new outcomes that the venture capital industry hasn't fully priced in yet.
The next unicorn may have only one employee. The next multi-billion dollar company may have no more than ten employees.
If you're a creator, operator, marketer, or founder who's been waiting for your license to enter the market, now it's here. Technological bottlenecks are disappearing, and startup costs are collapsing. What now stands between you and a real company isn't an engineering team, funding resources, or organizational size, but rather three questions: Do you have something worth expressing? Do you have the taste to judge good from bad? Do you have the discipline to deliver consistently?
This is the worst era in history for those who build homogenous software.
For those building products that are sharp, well-distributed, have communities, abundant resources, and in-depth expertise, this is the best of times in history.
Both of these things are true simultaneously. Which one applies to you depends on what you do over the next 18 months.
From here, the founders generally have three paths.
First, treat it as noise.
Convincing yourself that Apple is too big to fail, that your SaaS is unique enough, that AI programming agents are overhyped, and that everything will eventually return to normal—you'll have many like-minded people, because most founders will choose this path. And most founders will lose this cycle because of it.
The second point is to fall into panic.
The sudden shortening of the runway, hasty layoffs, and blind transformation are the price of reacting too late. Those truly destroyed by this transformation are not necessarily those who didn't see the changes at all, but rather those who only saw them 12 months too late and were forced to hastily change course without funds, time, or leverage.
Thirdly, take this 18-month window seriously.
Honestly examine your own moat, start building distribution pipelines before you really need them, find the differences that AI cannot replicate, and prepare for the coming world, rather than continuing to optimize for the old world you want to preserve.
Navarre's statement was very restrained yet very clear: "Pure software isn't worth voting for."
This isn't the opinion of someone seeking to avoid risk, but rather the final conclusion of someone who has spent twenty years judging what's worth investing in and now believes that most of what's currently being invested in is no longer worth investing in.
Apple has entered a period of structural death, and most SaaS founders may be next. Those companies that ultimately survive will be those that, upon hearing this assessment, begin taking action before everyone else realizes it.
The window is open, but it won't stay open forever. The real question is: in the next 18 months, are you building a moat that can withstand the test of time, or are you watching helplessly as the existing moat crumbles under the weight of reality?
Most people won't make it. A few will. The difference lies in what you start doing this quarter.
Related reports
Replit CEO: AI has sounded the death knell for SaaS; the next "Satoshi Nakamoto" is on his way.
