Let me tell you a quick story that perfectly captures our weird, complicated relationship with AI right now.
Imagine you're a senior consultant, a veteran with decades of experience. You're handed a massive pile of work—over 1,000 business requirements that need to be validated. It’s the kind of project that usually takes weeks of painstaking effort. But this time, it’s already done.
You’re told a group of bright, junior interns fresh out of school completed the analysis. You review their work, and honestly, you’re impressed. It’s detailed, insightful, and incredibly thorough. You give it a top-notch score: about 95% accurate. Great job, kids.
Now, picture a parallel universe. Another senior consultant gets the exact same completed project. The same insights, the same analysis, the same answers. But they’re told something different: the work was done by an AI.
What do you think their reaction was? They rejected almost all of it.
This isn't a hypothetical. It was a real, quiet experiment run by SAP to see how their consultants felt about their new AI co-pilot, Joule. It wasn’t until the skeptical team was forced to go back and check each answer, one by one, that they realized their mistake. The AI’s work was, in fact, incredibly accurate. The final score? Yep, about 95%.
So, What's Really Going On Here?
This little experiment is so revealing, isn't it? It shines a huge spotlight on the human side of AI adoption. It’s not about the tech; it’s about trust.
Guillermo B. Vazquez Mendez, a chief architect at SAP, put it perfectly. “The lesson learned here is that we need to be very cautious as we introduce AI,” he said. It’s all about how we talk about it and, more importantly, how we integrate it into the workflows of people who have built careers on their own expertise.
Let's be real—the resistance makes sense. If you’ve spent 20 or 30 years honing your skills, building a deep well of institutional knowledge, you’re not going to blindly trust a black box. There’s an understandable, and probably healthy, dose of caution.
But here’s the thing that often gets lost in the "AI is coming for our jobs" panic: tools like Joule aren't designed to replace that hard-won expertise. They’re meant to amplify it.
AI Isn't Here to Steal Your Job, It's Here to Do Your Chores
Think of it this way. An AI copilot isn't the new hotshot trying to take your corner office. It’s the super-efficient assistant who preps all your files, highlights the most important documents, and handles all the tedious grunt work, so you can walk into the big meeting and do what you do best: strategize, connect, and solve problems.
“What Joule really does is make their very expensive time far more effective,” Vazquez Mendez explains. “It removes the clerical work, so they can focus on turning out high-quality answers in a fraction of the time.”
This is the message he’s constantly repeating: AI is a tool for you, not a replacement of you. Human oversight is always, always part of the equation. The difference is, instead of spending your day digging through documentation, you’re getting that time back to think bigger.
Historically, a consultant might spend 80% of their time just trying to understand the technical guts of a system—how data flows, how processes run. Meanwhile, the customer is spending 80% of their time thinking about their actual business. See the mismatch?
“There’s a gap there—and the bridge is AI,” Vazquez Mendez says. AI handles the heavy technical lifting, which flips the script. It allows the consultant to spend far more of their energy on what truly matters: understanding the customer’s world and helping them succeed.
A Surprising Side Effect: Helping Newbies and Vets Connect
Here’s another cool thing that started happening. AI is actually changing how new people get up to speed.
We all know the feeling of being the new person on the team. You're trying to learn the ropes but don't want to constantly bug the senior folks with basic questions. This is where an AI copilot becomes a game-changer for junior consultants. They can use Joule to find answers and operate more independently, which helps them learn and contribute way faster.
But it also creates this amazing synergy. The senior experts see the new hires adapting and learning at an incredible pace with AI. That momentum gets them curious. It shows them, in a very practical way, what the tool is capable of and encourages them to start using it themselves.
This is also where they learn the new language of AI: prompt engineering. A huge part of the work now is learning how to ask the AI the right questions. For example, you don't just ask for an analysis. You tell it:
- "Act as a senior chief technology architect specializing in finance and SAP S/4HANA 2023."
- "Now, analyze these business requirements."
- "Deliver the output as a set of tables and a PowerPoint presentation."
Once people get the hang of framing prompts like this, the quality of the answers they get back skyrockets. It's a new skill, for sure, but one that pays off immediately.
Where We Go From Here: From "Toddlers" to Thinking Partners
As powerful as these tools are, it’s important to remember where we are on the timeline. “We’re still in the baby steps of AI—we’re toddlers,” Vazquez Mendez admits. Right now, the quality of the output is heavily dependent on the quality of the prompt. You have to guide it carefully.
But this is just the first phase.
The next leap is into what people are calling "agentic AI." Imagine an AI that doesn't just wait for your command. Instead, it can look at an entire business process—all the steps, all the dependencies—and start to reason about it. It could identify where a human needs to step in, or spot a task that an AI agent could handle on its own.
This is where a company like SAP has a massive head start. They’ve spent the last 50 years mapping over 3,500 business processes across pretty much every industry you can think of. Their systems support something like $7.3 trillion in global commerce every single day. That’s an unbelievable foundation of data and process knowledge for an AI to learn from.
With that kind of insight, the AI can move beyond just answering questions and start helping to solve complex, multi-step challenges. It’s a future where AI becomes less of a simple copilot and more of a true thinking partner.
And it all starts with overcoming that initial, very human, gut reaction we saw in that first experiment. It's about learning to trust the tool, not as a rival, but as a powerful new way to make our own expertise even more valuable.
