Let’s talk about the stuff that keeps me up at night, shall we? It’s a little thing called Service as a Software, and it’s incredible and terrifying all at once.
I was lucky to be part of the Software as a Service (the other SaaS) boom in the early 2000’s. Watching previously manual tasks become automated made so many professionals’ lives so much easier. Software sold to people for a monthly fee massively increased productivity. But many people at that time worried what would become of the workers who had previously done some of these more manual tasks? What would happen to the people who input and stored client and prospect information after Salesforce rolled into town? What would happen to IT team that oversaw the company servers and data centers now that Dropbox was available?
Many people panicked as software tools came in and took these menial tasks for people. But the net effect of the Software as a Service boom was an increase in productivity for workers and massive paydays for those SaaS companies that made it big. While some jobs did go away, employment in companies that used these tools actually grew.
Fast-forward to today, and once again the world is shifting thanks to technology, whether we like it or not. Thanks to AI and better large language models (LLMs) the era of Software as a Service has completely flipped on its head.
Whereas Software as a Service provided tools to workers, the new paradigm moves the needle even further: replacing the person altogether.
We are now entering the era of Service as a Software.
What is Service as a Software?
Service as a Software is software applications that provide services leveraging AI agents and systems of agents to automate tasks, make decisions, and interact with users. These services are designed to be scalable, flexible, and adaptive, and unlike Software as a Service, these AI agents can respond to changing user needs and environments. They can also be tailored to specific industries, jobs, or clients. In essence, Service as a Software doesn’t act as a tool to help workers, but rather more like the worker themselves.
And in some people’s minds, Service as a Software means replacing workers entirely.
How AI Agents are Replacing Human Workers
Service as a Software has significant implications for human workers. AI agents, specialized LLM AIs trained to do specific tasks in a way similar to humans, will become increasingly sophisticated and perform tasks that were previously the exclusive domain of humans. Need someone to do customer support? AI can do that. Need someone to reply to inbound leads and verify their intent to buy before sending to sales? AI can do that. Need in-depth analysis of trends happening in your industry. Guess, what? AI can do that. The list goes on and on.
According to McKinsey, 30% of hours currently worked across the US could be automated by 2030, with most of this transition happening in office support, customer service and sales, food service, and production work.
A key benefit of Service as a Software is that it’s quick to scale and its returns compound over time. Unlike hiring humans, once you’ve trained an AI agent once, you can easily replicate it over and over, and each agent will be instantly just as skilled as the original. There is no need to train them again, unlike hiring people.
On top of that, groups of AI agents will also be put together to create systems of agents, which will be able to talk to each other and combine complex ideas from multiple sources to solve larger and larger problems.
It is these systems of agents that I think will create the biggest disruptors in the economy and where Service as a Software will start making inroads into human jobs at a greater rate than what McKinsey believes. In the above graphic, McKinsey sees positive growth for most industries, but I fear they are underestimating what systems of AI agents will be able to do.
If a system of agents focused on law gets to the point where it is 99% accurate (AI has already passed the Bar Exam, btw), what law firm is going to keep a large, expensive staff on hand? If a system of agents can analyze the productivity of workers, create schedules and help with conflicts between employees, would the role of managers be less and less needed? AI is already to the point where it can code better than an average developer, wouldn’t those roles be slowly eliminated?
Service as a Software has the potential to completely upend our economy, slowly alleviating businesses of their biggest expense: people.
Will Service as a Software Replace Humans?
If left unchecked, businesses will do everything they can to maximize revenue and decrease costs. With human capital being the biggest cost for most businesses, it makes sense that the rise of Service as a Software would result in massive amounts of jobs lost. The stat about 30% of work being automated is a huge amount of human work that is suddenly not needed. Just imagine if ⅓ of the jobs in America disappeared. What would happen?
So is there hope?
Well, I hope so.
Just as technology has replaced people in the past, humans learn to adapt and move on to bigger and better things. In 1923, coal mining employed 883,000. Today it employs 43,000.3 You also probably don’t know many ice cutters, or elevator operators.
New jobs will be formed and done by humans, at least for a little while, during this transformative time in our economy.
And many jobs will become only partly automated. Mundane tasks that are easily replaced by AI, will create more time for humans to spend thinking critically or being creative in their jobs instead of being bogged down in those tasks. This could be a major net-positive for workforces.
The Responsibility of Leadership in the AI World
Automating these roles with service as a software isn’t just a way to cut the budget. It’s an opportunity to build up an entire workforce. AI agents can augment existing workforces in a way that makes them more effective, less frustrated, and more creative. But that requires proper training, freedom to explore these tools, and patience from business leaders to allow their teams time to adapt to this new landscape that is changing faster than ever.
Service as a Software Should Create, not Reduce New Jobs
As Service as a Software becomes more widespread, businesses are being forced to adapt to this new reality, and many will and are acting rashly. Many companies hear about their competitors using AI to cut costs, and want to do the same. The result is a focus on replacing jobs with AI to decrease the cost of human capital and increase profit margins.
What many organizations are completely missing is that Service as a Software requires us to re-analyze traditional jobs at our companies. Instead of looking at AI to replace entire jobs, look at what jobs are out there and break them up into the individual tasks that make up the job. Then, rebuild those jobs to be in alignment with the vision and mission of the company.
In all of my years as a business leader, one thing I’d hear over and over again was “I would love to do that, but I just don’t have time.” It was an ongoing theme of employees that felt burnt out from trying to just keep up with the grind of work. They all wanted to go above and beyond what they were doing, but there were so many little tasks that had to be done that they couldn’t do the things they really wanted. So what if we re-analyzed the jobs we have in place at our organizations and thought about the jobs we would need to make everyone else less burnt out?
At Ceralytics, we hated pulling the data for the reports we presented to clients. It was mind numbing work. So we created a solution that pulled the reports for us, giving the team more time to really think about solutions for clients instead of worrying if they properly copied and pasted a number correctly from Google Analytics. This “report puller” could have easily been a person’s job, but we gave that job to a Service as a Software, enabling it to pull the reports for us. No one lost a job as a result of that change. Instead, it made everyone happier and more productive.
We didn’t use Service as a Software to replace a person. We used it to fill a new job at our organization.
The Big Shift Service as a Software Brings
Many job roles are defined by the knowledge the person in that position has. For sales, it means understanding the benefits of the product and how to convey those benefits to an end users. Customer support, on the other hand, required some solid knowledge of the product but did rely heavily on automated scripts for many calls.
In the age of Service as a Software, these roles suddenly blur together. Could sales and customer support blend together utilizing an in-house “expert” AI to answer questions in real time for a client or potential client? If the expertise lives 18 inches in front of the user, wouldn’t the human skill of connecting with someone and helping them figure out a problem be the new role of sales and customer support? You could have the same number of people as before, but now they can not only sell the product, but help clients and upsell them when it makes sense. The only change you’ve really made is taken the expertise and made it an AI agent that assists people.
Think about all of the mindless tasks you face every day. How much more effective would you be if you had a personal assistant who could take care of 30% of your work for you? My guess is you’d be more than 30% more effective. And that’s the point. The value we get back out of these systems isn’t just what they can do, it’s what gift they give back to us.
And that gift is more time to be human. More time to really connect with other humans.
If business leaders can embrace AI and use it to offset the tasks that bog people down, those employees will have more time to use their creativity and critical thinking skills to do their jobs even better. This should be the future we’re trying to build. We shouldn’t be looking at Service as a Software as a means to get rid of jobs, but as a way to automate new jobs that make everyone else’s jobs more effective.