Understanding the potential of co-pilot AI

Nov 28, 2023
What is Co-Pilot AI?

Let's begin by painting a picture of the world today. In every industry, knowledge workers spend a significant amount of their time on repetitive digital tasks. These tasks can range from sorting through emails and scheduling appointments to updating spreadsheets. Although necessary, these tasks consume valuable time and energy that could be better spent on more meaningful and productive work.

This is where co-pilot AI comes in. Co-pilot AI tools, such as Cykel, act as digital assistants that can perform these mundane tasks for you based on natural language instructions. Just like a co-pilot in an airplane assists the pilot by sharing the workload, a co-pilot AI shares the workload with you, handling the routine tasks while you focus on the bigger picture. By doing so, it empowers you to be more productive and use your skills and talents in ways that truly matter.

Software has delivered huge benefits and productivity gains for society. Before we had paper ledgers, encyclopedias and physical maps, then software brought us spreadsheets, Wikipedia and Google Maps. So we've seen the world become more efficient as software empowers us to do more with our time. However, we're now at a turning point where computers can perform tasks for us by interacting with software on our behalf.

Today, a layer of user interface sits between our intent i.e. what we want the computer to do for us, and the end result. We go to a website and interact with a series of visual metaphors – a keyboard which looks like a typewriter, a text box that looks like a paper form, a switch that looks like a light switch, and so on.

This abstraction has been a helpful solution but, with the advent of sufficiently intelligent AI models, we are now at a moment where software can fully take over and complete tasks on our behalf, reducing or even removing the need for us to interact with user interfaces.

The user interfaces we're accustomed to—buttons, forms, switches—were designed to make the interface with digital systems easier. However, they still require time and effort on our part. With co-pilot AI, we're decoupling from these constraints. Instead of interacting with a series of user interface components, you simply communicate your need through natural language, and the AI assistant understands and executes the task for you.

This has the potential to significantly change the way we work, unlocking hours that can be repurposed for more meaningful, impactful work. It's not just about efficiency, it's about using computers in a new and improved way.

AI that can interact with anything on the internet

People have spent the last 30 years building software systems which are able to perform actions in the world - booking a flight, finding a house, getting information about a news event.

With the internet and software systems, we've built a vast digital infrastructure connecting the world together. The CRUD database—standing for create, read, update, delete — is the backbone of these systems, allowing us to manipulate data in helpful ways. When you break it down, any software system is a user interface on top of a CRUD database.

The traditional user interface is a neat solution to the problem of how to interact with a database but also creates a barrier, requiring us to manually input information and navigate through various steps to achieve our desired outcome.

With co-pilot AI, we are removing this barrier. Instead of having to manually interact with a series of user interfaces, you can simply tell your AI assistant what you need, and it will navigate the digital infrastructure on your behalf to complete the task. The AI can chart the course and steer the ship, freeing you up to focus on your destination.

Co-pilot AI systems are designed to harness the power of the digital infrastructure we've built and use it to do useful tasks. This is a foundational shift in our relationship with technology, moving from a manual, hands-on approach to a partnership between two intelligent systems.

Single-platform AI features vs cross-platform AI co-pilots

Recent advances in natural language processing (the technology that allows machines to understand and process human language) have made it possible for AI assistants to not only understand the tasks we ask of them, but also to execute these tasks with a high degree of accuracy and adaptability.

Cykel uses these breakthroughs to power our co-pilot. You can simply tell Cykel what you need, and it will go to work for you, interacting with the necessary software systems to complete the task.

What about just building AI into the software systems we use? With the increase availability of AI models, many products are now incorporating AI into their UX. You may have noticed that these products now have generative AI features that allow you to compose emails or summarise text. While there is some amazing work being done, the benefit of co-pilot AI is that it isn't confined to just one system. It can perform complex tasks across multiple systems and platforms.

While a platform implementing AI can improve one workflow within that platform, co-pilot AI systems can perform tasks across platforms, regardless of whether the platform itself has implemented AI features. With Cykel, you can move from one website to another understanding the context, background about the user and what they're trying to do. You can take that information and perform complex, multi-step tasks that require navigating through different platforms and systems, all while maintaining context.

This also means you can seamlessly integrate into a user's existing workflow and systems. Users don't have to waste time and resources on setting up and learning a new tool, instead, Cykel works with the systems and processes they already have in place.

“Human-in-the-loop” AI

As with any technology, there are limits to what co-pilot AI can do. There are still challenges and limitations we are working to address. It's important for users to have realistic expectations and understand that co-pilot AI is a human-in-the-loop tool to aid in productivity, not a replacement for human creativity, judgment, and decision-making.

Human-in-the-loop systems represent a crucial intermediate step in the evolution of AI. While there is significant research and development focused on creating fully autonomous AI that can fully replace human input, the reality is that we are not there yet. Instead, what we have are systems that greatly enhance human capabilities and automate some percentage of their work.

With human-in-the-loop systems, the AI and the human work together, each bringing their unique strengths to the table. The AI can handle the mundane, repetitive tasks, process vast amounts of data, and even make recommendations based on that data. However, the human is still very much in the driver's seat, providing context, judgment, and creativity. This collaboration allows for more efficient and effective task completion, as the relative strengths of the AI and human are used more efficiently.

Getting started with co-pilot AI

You can learn more about Cykel Co-Pilot or read about our company and mission.