Welcome to the era of the AI co-pilot

NBC’s The Good Place is set in a heaven-like afterlife complete with whatever the show’s lead characters can dream up to make the rest of all eternity as enjoyable as possible. Think flying lessons and endless amounts of frozen yoghurt in every conceivable flavour.

Oh and Janet, of course. Janet is a super smart personal assistant who appears out of nowhere and can instantly conjure up any requested object. In addition, Janet is also a walking database of all the information in the universe; with a wide understanding of every event that has ever happened on earth to anyone who has ever existed.

The more I use them, the more I feel like AI co-pilots are our very own Janet. This new generation of AI tools is capable of making our jobs easier and enhancing our lives with a single request.

Capable of producing results akin to magic, these intelligent virtual assistants leverage large language models (LLMs) to understand, analyse, and process vast amounts of data.

In action, AI co-pilots can automate repetitive tasks and respond to user needs based on contextual information. They can also quickly make sense of large amounts of data; identifying patterns and trends and bringing together information from different tools, apps and platforms so that you have everything you need to make the best possible decisions.

As an example of this, our software development team recently created a Slack bot called Claude. Claude is a large language model that is really good at summarising Slack conversations. If a member of the team was in a meeting and missed a lengthy discussion on Slack, Claude will pull out the most salient points and package them in a way that is easy to follow and understand.

Similarly, Optimizely has incorporated an AI assistant into their CMP platform to help marketers write superior content, generate social snippets or put together better web copy quickly and easily. This AI assistant aims to streamline time-consuming editorial activities so that copywriters can work smarter, not harder.

A bot for everything

Given the time and cost savings associated with co-pilots, surely it makes sense to have these nifty virtual assistants for everything? Well, not exactly. As is the case with any new technology, it’s important to ask why you’re making the investment and if the solution you’re deploying is really suited to your needs.

Just because everyone is developing a chatbot, doesn’t necessarily mean that you need to bring a conversational experience into the mix too.

This is particularly true if it leaves users feeling confused as to why a conversational tool exists at all. When writer, Scott Nover, decided to try out a few different bots, and found himself asking this exact question.

AllTrails is a global platform that showcases the best outdoor trails and nature experiences. Their bot promises to help outdoorsy folk find trails for their next “hike, ride, or run.”

While the idea of a hiking, running and biking route recommendation bot isn’t overly offensive, Nover questioned if it would have made more sense for AllTrails to rather put together an interactive map that would easily allow users to browse for routes in their area without having to have a conversation with a bot.

So, how does one figure out how to use co-pilots effectively? Here are a few pointers.

Choose the right solution: With so many different AI co-pilots out there, it’s important to pick the solution that best fits your needs and is designed to assist with the type of work you do. Start by defining what you want your copilot to deliver. This will determine what kind of product will be most suited to what you want to build.

Start small… and simple: Remember that a copilot is a probability engine, so the more broad or vague the request, the higher the likelihood that it will produce random results. As such, avoid trying to get your AI co-pilot to do everything all at once. Start with simple tasks before challenging it with more complex requests.

Master the power of prompting: Just like a good teammate, a copilot needs clear instructions in order to shine. The more specific your prompt, the more relevant the result. Here, it’s important to play around a bit and test out different prompts so that you can see what works best. Remember that these tools are always learning so the more a copilot is used, the better equipped it is to understand your preferences.

Review, review, review: I cannot stress this enough – the output produced by any AI tool should not be seen as gospel. AI co-pilots are still under development, so you need to make sure that the results are accurate and appropriate before using them. It is essential to triple check all content generated by AI to ensure that it meets your needs.

In a nutshell, AI co-pilots simplify complex tasks, provide guidance and support, and have the potential to totally redefine how businesses operate in the years to come.

Microsoft’s executive chairman and CEO, Satya Nadella, describes co-pilots as a technology that will usher in a new era of personal computing. Helping us to gain access to the world’s knowledge and, more importantly, empowering us to act on this knowledge. As such, it won’t be long before co-pilots are a mainstay of everyday life.

Need help leveraging co-pilots in your business? Reach out to us, here.