In tech disruptions, the real game-changer isn’t always the invention itself but how it’s used. Take Uber and Airbnb, for example. These aren’t just apps; they’re entire ecosystems that reshaped how we travel and stay, built upon the foundational platform of Apple’s App Store.
Now, we’re at the brink of a similar revolution, this time in artificial intelligence, led by OpenAI’s closed-source GPT Store and challenged by Meta’s open-source Llama2.
Meta hasn’t yet announced that they’ll follow Google’s genius playbook of positioning itself as an open-source platform, i.e., Android’s Play Store, competing against the closed-source Apple App Store. The main thesis of this article is that there is a short window where Meta can seize the opportunity using Llama 2’s striking success to position itselves as an open foundational AI platform and GPT marketplace that would dwarf the smartphone app store revenue potential.
But let’s go step by step.
Is OpenAI’s GPT Store the New Apple App Store?
Imagine an app that could plan and book your entire vacation in minutes, a GPT-based real estate broker who not only researches potential buildings to buy or rent but also negotiates on your behalf, or a personal shopping assistant that knows your taste better than you do.
Thanks to OpenAI’s platform, the GPT Store, these are no longer fantasies. It’s the AI equivalent of the App Store, a closed ecosystem ripe for developers to create applications that could potentially outgrow the platform’s revenue and impact.
“A platform is when the economic value of everybody that uses it, exceeds the value of the company that creates it. Then it’s a platform.” — Bill Gates
💡 A sustainable and healthy platform is one where the apps or services built on it are worth more together than the platform itself. For example, in Ethereum, the value of all its apps should be higher than Ethereum’s own value. The same goes for Apple’s App Store, where big companies like Facebook, Airbnb, and Uber make more money than the App Store. This shows that the platform is good at helping new ideas grow and succeed.
OpenAI’s GPT Store, like Apple’s App Store, is closed-source — nobody can change, modify, or clone the platform itself — because it helps Apple keep high-quality standards and control over their technology, which leads to big profits. Being closed-source means they can decide who uses their tech and how making sure everything works well and stays safe.
The biggest closed-source platform like the App Store or GPT Store usually comes with a strong network of users and developers that are interconnected on the same platform. It’s hard for new competitors to break into this network, as users and developers are already used to and trust the existing system, making the business stable and hard to disrupt.
The biggest closed-source platform also has a strong data advantage — data is the only real long-term differentiator or “business moat” in the AI race.
However, there is one way to break these moats, and we’ve seen the very familiar playbook in the smartphone world during the last two decades. Read on to learn the whole story…👇
Meta Will Come Back — If They Seize the Opportunity to Position Llama2 as the Open-Source Contender of OpenAI’s Closed GPT Store
There’s a big chance now in the AI world for a more open way of doing things. Companies like OpenAI have their own closed AI systems where developers can make apps. But what’s missing is a more open option, a platform and foundational AI model that is free to use, open-source, and change by anyone.
It is still unclear who will fill this gap. Some new players are Anthropic with Claude 2, some scientific collaborations for AI training, and blockchain-based AI systems that are open and “decentralized” — although all I have seen are DINO, i.e., decentralized in name only.
Then there’s the 600-pound gorilla in the open-source LLM space: Meta with Llama 2 that is positioned nicely to be able to offer an open AI platform similar to the Android opne-source app ecosystem. Will they seize this one-time opportunity?
As history shows, closed systems often invite open-source competition. Meta’s Lama2 might be stepping up as the Android to OpenAI’s Apple. It offers a more accessible platform for developers, potentially leading to a broader, more diverse range of applications. While it might not boast the same profitability as OpenAI’s model, its impact and adoption rate could be staggering.
This is exciting because it can make AI more available to more people, helping everyone to use and make AI stuff easily.
The MMLU Benchmark performance chart clearly shows that open-source AI models show an even faster slope of performance improvements as compared to their closed-source contenders:
Here’s the App Store Revenue which approaches the $100 billion per year mark:
For comparison, the mobile ad-spent — a $300 billion per year revenue business!
Investing in the Future: The Meta Perspective
This brings us to a crucial investment insight.
Just as Google’s Android carved out a massive slice of the market, Meta’s open-source approach with Lama2 could position it as one of the largest player in the AI app economy. For investors, this means considering Meta for its current value and potential to lead the next wave of tech innovation.
The leadership in open-source LLMs may be one of the reasons the Meta stock has shown a dramatic rise during the last 12 months?
However, look at the Meta revenue chart — doesn’t look pretty because the long-term growth of revenue from social network ads is broken:
So, while there may be AI platform revenue in the future, the revenue from ads is not as stable and Meta needs to act decisively and quickly to create the Android for OpenAI’s Apple app store.
The New Wave of AI Applications
In any case, the potential applications of these AI platforms are limitless.
Much like almost every business distributes its own app on the Apple and Android app store, every serious business will build on the OpenAI GPT store. Only the revenue potential will be much higher. We’re talking about scaling intelligent labor!
The TAM for scaling intelligent labor is limitless.
From AI-driven content creation tools that could revolutionize marketing to fixing the health system through advanced healthcare diagnostics that are orders of magnitude better — patients could be monitored 24/7 through AI doctors instead of the 5 minutes per year care we’re currently seeing by physical doctors.
These applications, built on platforms like GPT or Lama2, could dwarf their parent platforms’ revenue and societal impact.
We’re not just witnessing the growth of new technologies; we’re seeing the birth of new economies. Platforms like OpenAI’s GPT and Meta’s Lama2 are the fertile ground from which the next Uber or Airbnb of the AI world will grow.
For tech enthusiasts and investors alike, understanding and leveraging this shift is key to being on the right side of change.
🧑💻 TLDR: The next wave of technological disruption is likely to happen on top of one or two large GPT stores. There will be significant revenue opportunities for builders on top of these foundation models. There will be significant opportunity for investors spotting these trends early. Think Google and Apple in the early days!
Use GPTs to create value, join Finxter, buy scarce assets like Bitcoin, invest in AI tech companies, be on the right side of change. No financial advice. 😉
💡 Recommended: Prompt Engineering with Llama 2 (Full Course)
Jean is a tech enthusiast with a love for AI and machine learning innovations, particularly LLMs. Beyond contributing insightful articles to our blog, Jean has worked as a Python, Rust, and Go coder for one of the leading tech firms in the world.