5 things to consider so your AI adoption doesn't fail you

General Motors has just laid off ~600 salaried IT employees (about 10% of the department). The company's general hiring trajectory is indicative of a larger shift: It's obvious by now that global enterprises are mostly interested in people who know how to work with AI-native workflows.

There's a backlash on X, there are discussions on Reddit, and people are even quoting the fresh encyclical by Pope Leo XIV to make sense of the layoffs. In the past year, we've seen more companies acting like they're forced to restructure their whole workforce to foreground AI. Whether you agree with the Pope on the matters of AI's influence on the workforce or not, major changes in the hiring trajectories of large enterprises generate the question, 'Why didn't these companies build AI capability into their teams in the first place?' General Motors has the capacity to make such drastic decisions on the go (even if the process lacks certain elegance), while middle-sized companies don't.

Here are five things business leaders need to know before they're forced into a crude transition.

Read more:  New workers are missing out on 401(k) savings, and employers can help

AI won't fix every problem

One of the biggest mistakes businesses make is adopting a range of AI tools and expecting it to fix all of the company's inefficiencies. For quite some time, AI indeed was presented to the general public as an efficiency tool. Well, it's safe to say we're past the point where AI agents are there to solely "boost someone's productivity." For starters, to work well, AI — and especially autonomous agents — need a robust, well-prepared infrastructure of processes and data to lean on. AI as a productivity tool might provide a small boost, but if you haven't optimized the process it's part of first, you won't see the results you're expecting. At the same time, if a process is already standard, clear and does not require an LLM to carry it out, there is no need to use one.

Businesses need to practice a problem-first — not an AI-first — approach. Break the process down, pinpoint the real issues and select the best-fitting solution. That might be AI, or it might be human — your goal shouldn't be to implement AI for everything, but to figure out where it's really going to serve your business.

AI adoption doesn't necessarily indicate meaningful progress

You can probably remember how the fear of an endless lockdown during the COVID pandemic made businesses pour money into the metaverse. It yielded some companies some traction and investor attention; however, for most, it was a waste of resources. Right now, we're witnessing companies making a similar mistake: Successful footwear brands and social media platforms are rebranding as AI companies to jump on the bandwagon.

The question is this: If all you have to show in terms of AI adoption results are press mentions and internal excitement, what have you really achieved?

In the long run, the winners will be those who apply AI where it really offers value. This means the value expressed in amounts of actual money, not only PR publications and social media engagement numbers. Deploying AI anywhere and everywhere in the hope that it will somehow raise your valuation is very tempting, but you have to intentionally plan for AI to help your company grow in terms of cost reduction, improved processes, higher customer satisfaction, and measurable productivity upgrades.

Read more:  How companies can keep their lone workers safe

Redesigning roles can't happen without employee's input

Much like during the pandemic, the tensions in the market are high right now. People are understandably concerned for the future of their own position and their professions in general. Insisting on AI adoption where it isn't valued won't help you keep the employees happy or make them work harder. So, before you make any changes, consult those who will be most affected when bringing in AI-first or otherwise AI-focused workflows.

Additionally, now is the perfect time to listen to your team leaders. They know best which processes cause headaches and make the job more difficult. Rather than adding complexity where it isn't needed or wanted, focusing on those issues is what will get employees on board. Redesigning roles and restructuring your teams is a step you can only take after the whole mapping-out-the-bottlenecks task is complete.

You're only as strong as your weakest link

People who work in retail and supply chain management know very well that overall productivity and speed are only meaningful when the whole chain is evenly productive and agile. If a supplier delays production or a logistics partner shuts down, so does the whole company. Surprisingly, the large-scale deployment of AI agents exposes the very same issue, just for more industries.

Many businesses aim for isolated productivity gains, and introducing AI agents does seem like it'll speed up some tasks. Say, your employees would spend less time coding. Now, if your company's main bottleneck is in getting the management approval (and not the coding itself), what will the five-minute win save you in the end? Instead of measuring AI success by what a single tool delivers, we need to start looking at the entire process. Is it actually getting your employees to the finish line faster? Is the end product any better? Are employees any less stressed?

Read more:  Employees aren't taking their time off. Can 'PTO-maxxing' help?

Understanding the shift

The businesses that navigated COVID best were those that accepted the world had changed. They didn't treat remote working as temporary, but invested in collaboration tools that offered flexibility moving forward. During COVID, some teams changed their workflows to async, remote-first work, while others simply transferred all their old office habits to the online environments and had a hard time.

With AI, businesses are now at the crossroads again. We can either change the way we work or stick with old workflows and get much less value from the technology. AI isn't some passing trend, and it's not going away anytime soon. One can't just stick to protecting old ways of working, because it would mean sacrificing a huge growth opportunity for comfort in a very short run. 

In the long run, this same comfort puts your workforce at far greater risk of job losses than accepting and adopting AI, after thoroughly considering what you want to reach with this adoption. The Tolkien quote chosen by the Pope is, after all, a very accurate description of the attitude we need "for the succour of those years wherein we are set."