The great software rewiring: AI isn’t just eating everything; it is everything

Once they once eaten the program the world. Now it’s here to digest what AI is left. The applications are managed, the input and platforms managed by market places, the old model, which was cut, said. What is something that arises AI-First World There are not trapped inside the applications applied within the programs, but dynamic, required services are available as access to the EU-native interfaces.

Computing for decades, checking the wardrobe. Applications were digital folders, self-owned, hard and walled. Want to check the weather? Open an application. Need to order flight? Another application. You have to pay a bill? Another one. The result? One fragmented user experience we switch between countless silos, each competes for real estate on a home screen.

Generative AI violates this model. Instead of clicking and shooting through individual programs, users will interact with the dynamic, dynamic, process and creator of the process, no application is required. Manage Travel from the TEK AI Assistant, optimize finances and recommend a training rule? Did. Do you need legal documents when ordering grocery products and generalize today’s news? Seamlessly. The new interface is not an application. This is a conversation, predictant and friction.

This new world of functional intelligence to be fair is still not completely ready. Applications do not disappear in one night but can slide the grip on the calculation. AI Previously packaged software does not care about silos. The program recovers the module, speaker and practice using Deep integrated. Opening between programs and all the opinion of the passage? It will quickly feel like inherited inheritance.

Available Risk: Traditional Sunday Places are in the hour

For years, the digital store and walled garden markets were incredibly moats. Distribution of control, taxes each operation and billions of climbing. Nice. But what happens when there are applications …. unnecessary?

The increase in AI-based communication is all threatening App Distribution Economy. If users rely on the AI-native systems instead of installing discrete program, the traditional software market markets are becoming a relic. Eat the AI ​​broker. The economic model is where the application monetinated is not in the EU’s overhauled service layers, unemployed, personalized and most importantly, inherited platform control.

Two inescapable results:

1. Income Violation: No 30% decrease in application sales or purchases in the application. If AI autonomies AI transactions, the App Store is the implode of the economy.
2. Platform Disintion: Ai Cloud-native and hardware-agnostic. Control over digital ecosystems, software is not an experience, but reduces the environmental service.

Who is the new question owns AI’s service layers? Because, who has the next trillion dollar industry.

New power structures: AI models and vertical AI solutions

AI cooking application creates an open electricity vacuum. Where does the value change? Simple, control:

• AI models: The most advanced basic models are determined by the basic intelligence layer.
• User Interface and Personalization: Whoever has the most intuitive, EU-native interfaces, the sign will be dominated.
• Information and Integration: AI allows you to access real-time, property data. Anyone who owns the information pipelines manages the concepts, intelligence and the economy.

But there is another force in the game: vertical AI solutions.

Currently, the largest language models (LLMS) feel like a Swiss army blade with endless tools – exciting, but excessive. Users do not want a “understanding” AI. They want AI agents that are suitable for special industry and work flows. Think: Legal AI project contracts, financial AI management investment, creative AI generating content, scientific AI accelerating research. Broad AI is interesting. Vertical Ai is valuable.

Currently, the LLMS is very wide, very abstract, very inaccessible for most. The empty conversation box is a product, not a product. If the AI ​​applications replace, users should be integrated into a problem without making you think of daily work flows without making you think about settings, parameters or backgrounds.

Companies who succeed in this next wave will not build better AI models, but better AI experiments. The calculation is not about an AI who does everything to the future. This is a very specialized AI systems that knows what users need and this perfect work.

The whole program stack is rewritten in real time. What does the old model replace?

• Microservices over programs: Forget Blooded Apps. Future software will be modular, required and Ai-Callable. To order travel? AI Agent, real-time, hotels and rental cars, before opening any application.
• AI-Powered Marketplaces: The next software market is not an app store. This is a AI-Native services market where the users are subscribed to functional AI agents, which users are on the function of function.
• AI-AS-A-Service: Instead of selling related applications, developers will develop “skills” or “agents”, which are monitoring through subscription or use-based prices.

Inevitable breach

This is not evolution; It is a blow. Gen AI is not just another layer of technology; All software industry has the potential to eat from inside.

The old program model was built on scarcity. Control distribution, restrict access, download fees. AI destroys it. The new model is liquid, frictionless and endless expanding.

Platforms and jobs, which cannot be adapted, can be well sent to the history books joining the internet, mobile and clouds.

AI is not just the next program wave; Wave that breaks everything previous. The only question is left: Who walks him and who drowns?

Justin Westcott leads the global technology sector Sketch.