“Finding the killer app?” should be “Making more millionaires?”

Many hardware companies are looking for killer apps. From Apple with the Vision Pro, OpenRAN networking, IoT Gateways, Smart Vacuum Robot, to AI PCs. If only that magical killer app would come along and now millions of people wanted to buy their hardware for it. “What is the killer app?” is actually the wrong question. Nobody bought the early iPhone because it had that one killer app everybody wanted. As soon as there was enough activity in the App Store, everybody had their own collection of apps they liked. Yes of course a lot today download WhatsApp, TikTok,... But there was not a single “killer app” that sold the iPhone to millions.

Now let’s put ourselves into the heads of a person that could potentially make a “Killer App”. Let’s say the “killer app” is to take images from a camera and to see if there is smoke or fire. When it is detected, the owner of the device gets immediately warned. The Vision Pro, an office networking device with a camera, an industrial IoT gateway with a camera, a vacuum robot, an AI PC,... would all be able to get the “Fire Detection App” and save lives. On a Vision Pro you would have to make apps according to Apple, based on their AI silicon and pay them 30%. The networking device is using ARM and some Chinese NPU. The Industrial IoT gateway is using AMD and Nvidia Jetson. The Vacuum Robot is using a RISC-V processor and some custom AI FPGA. The AI PC has an Intel processor with AI capabilities in it. So creating one solution is not possible today.

Imagine we would be able to do reality distortion. Assuming there is some open source standard to package the AI code and abstract both the camera image stream as well as the on-device AI accelerators, we could actually create the app. Luckily this is a solved problem technically but still not enough people know about it. Webassembly or WASM and some of its surrounding innovations like WASI can solve this issue.

So we have our killer app. We are able to connect it dynamically to the camera and the on-board AI accelerator. Now how do we interact with the customer and show them the smoke or fire. We could immediately make an alarm go off but if the house is full of people smoking cigars that would lead to false alarms. We might want to notify the user wherever they are. That could be a mobile notification. A TV screen notification. A PC desktop notification. A browser notification. A MacOS notification. Now you click the notification and see the images of smoke. Again making this a cross-platform experience requires some work. Google brought out Flutter which allows anybody to create apps that can be run on different devices from mobile, tablet, Web, PC, MAC, to potentially TVs,...  However quite some work would be needed for the developer to make a unified distribution solution.

The work needs to be done before any value is brought to the customers. Developers don’t like to work on something nobody ever will use. Ideally there is some type of place where you can easily find product market fit before you actually start coding. Find a way to validate that lots of people have a problem. Propose a solution. Even get pre-orders based on a prototype for a product that can be easily launched. Nothing like this exists. You have crowdfunding platforms but not customer problem focused platforms. It wouldn’t be hard to make the platform. Creating a simple way for companies to push a problem sourcing platform to their customers would be valuable. The platform can take a small fee and be self-sustainable and ideally is ran in a cooperative mutual style whereby nobody is the single owner.

The last and most important part is the financial incentive. Let’s say you create the Fire Detection app that can be installed on vacuum cleaners, IoT gateways, Vision Pros,... How are you going to earn money? Apple’s "30% goes to Apple" model is being rightfully challenged in courts and by the EU. More funds would have to go to the developer, e.g. the core store platform would get 5% and handle app formats, distribution and payment. An ecosystem of testing and security companies could validate the code is safe and get some percentage, e.g. 3%. The hardware vendor should get some percentage, e.g. 5%. The sales organisation selling the device and app should get 5%. Some SDK platform abstractions could get 2%, e.g. camera (0% because standard), notifications cross-platform (0% Flutter + price per 1,000 notification) and AI (2%). So at the end the developer gets 80% while 20% is paid to others for app store distribution/payment, device sales, hardware, base software, testing/security checking,... Some services have a usage paid fee, e.g. notification, cloud compute,..., which the developer would have to pay from the money that comes in. Via smart contracts and a DAO on the blockchain, this whole ecosystem could be made to work with revenue distribution being fully automated.

Conclusion

The main reason why we have not seen any killer apps beyond the mobile is that no other hardware ecosystem has been able to make millionaires out of the people who are able to create them. Too many hardware vendors think wrongfully they need walled gardens because that is what Apple did. Only few devices are sold at enough volume that they can warrant having a unique app model unfortunately.

Too many players think they deserve 30% because that is what Apple wrongfully thinks. Revenue needs to be based on value. The developer brings most value to one app so they should get the bulk of the value. Hardware, software and other platforms should get small percentages from thousands or even millions of apps.

We need to make it super easy to crowdsource problems, validate solutions, distribute solutions and make new millionaires, otherwise we are not going to see a lot of killer apps in the future.

I have a simple suggestion. If you are a hardware/silicon vendor, device sales organisation, SDK platform abstraction group, VC, device programmer,… that wants to launch killer apps and are unsure how to make more millionaires to get there, then reach out privately and tell me how much you are willing to sponsor towards an open source and cooperative platform that would make the above possible. Only if enough hardware/silicon vendors, device sales organisations, SDK platform abstraction groups,... come together, a broad standard platform could be made that gets managed by its members.

In Web3 there are tools to make these fully autonomous organisations whereby its members decide what the rules are of the platform and how any revenues are automatically distributed. These platforms can bring competitors together without creating cartels or making anybody less competitive, quite the opposite actually: everybody earns more. Greed of one type of member is a platform killer, not a killer app. Collective power is what changes things.

If there is no interest, this is just another blog post. If there is a lot of private interest, then hopefully this is just the first blog post about #MutualMarkets, the concept of open source cross-industry collaboration towards Mutually Owned Decentralised Marketplaces where all members win and there is no member who undeservingly gets most of the value.

If you want to be a customer, please use the comments and say what apps you would buy on what platform. Anybody else can help us win the war against the LinkedIn Algorithms by just clicking that Like or Share button 🙂