Good morning. Nearly all companies are using AI nowadays. So where does a real edge come from? Often, it’s the ability to tap knowledge that never makes it into a database, according to a recent report by Amazon Web Services.

“Every organization has access to the same AI models, so competitive advantage sits in their own data, and above all in what their people know that has never been written down anywhere a system could read it,” according to Tom Godden, an AWS executive in residence and former chief information officer.

Capturing and using that data is even more critical when deploying AI agents, according to Vikrant Bhan, Nestlé’s chief data and AI officer. “We were only looking at structured data,” Bhan said in the report. “We were not looking at all those [standard operating procedures] and factories and different documentation. What we have been exposed to in agentic AI is that world is actually, if not as important, even more important.”

It’s the same thing Bob Pisani, chief technology officer of financial software provider Addepar recently told me. While Addepar was figuring out how to integrate AI inside the company, its biggest challenge was applying “the knowledge that lives between the systems, between the people,” Pisani said.

That information can include the way someone works or does their job. “What I need is you to tell me how you do your job, how you think about decisioning, and how you make value judgments,” Pisani said. That’s the kind of “between the systems” information that needs to be written down so it can be ingested by AI.

In other words, it’s institutionalized or tribal knowledge, and it requires a lot of documentation to capture. But once that data—whether it sits in the day-to-day decisions that employees make inside their heads or within emails and Slack messages—is captured, it can be used to train and retrain an AI model in real-time.

“I call it the data breadcrumb trail, where really we’re saying, ‘Let’s look at all their actions from one user journey or one set of decisions over time,’ and then we put that together and we start to do lots of interesting things around combining machine learning and AI,” Pisani said.

The result? A “massive uplift” in accuracy and return from AI because Addepar was more efficient at farming the collective knowledge of its staff, Pisani said. The company has continued hiring engineers to supplement the additional productivity it has gained from AI tools, he added, and invests over $150 million each year in research and development.

• Companies are finding AI costs hard to predict. In a recent study only 11% of nearly 400 businesses surveyed were able to accurately forecast AI spending. “Unlike traditional software, AI behaves more like a human worker: It takes action, makes decisions, sometimes even makes mistakes—all on the clock,” the WSJ reports. Pricier AI models can end up cheaper overall because they handle tasks more efficiently, while cheap models given the wrong job can rack up tokens.

• A new White House task force, dubbed the “Super Intelligence Force,” has 120 days to report on the risks and opportunities of artificial intelligence and to define the federal government’s role, the WSJ reports. Leading the effort is Director of National Intelligence Jay Clayton, who effectively becomes the administration’s AI czar. Though the White House has mostly argued for lighter regulation, the task force suggests a growing acknowledgment that AI will require at least some federal oversight.

• Companies behind the rash of rogue AI agent episodes can also do better and it starts with culture, suggests David Robinson, a former OpenAI safety leader who recently left the startup. “Frontier labs need to run like nuclear-power plants or busy airports, with layers of redundancy and careful, time-consuming planning,” he writes in the Atlantic.

• German robotics startup RobCo is now valued at more than $1 billion, double its valuation from January, after a $40 million share sale mostly by employees to investors including Sequoia and Cherry Ventures. The deal reflects booming investor interest in physical AI, as RobCo pushes into the U.S. with Alfie, a self-learning, two-armed factory robot it pitches as a fix for manufacturing labor shortages.

• Deutsche Telekom expects AI and automation to generate about $1.24 billion in gross savings outside the U.S. by 2027, the WSJ reports. The German telecom group also says building the technology into its networks and European infrastructure will open up new customer services and revenue. Its projects include a platform that lets businesses automate recurring tasks and an AI assistant that translates phone calls in real time, which could lead to new services and rate plans.

The WSJ Tech Council brings together CIOs, CTOs and CISOs advancing innovation and shaping the future. Join this trusted community where tech executives connect with peers to explore emerging trends and gain the perspective they need to stay ahead of disruption.

Request Information

Follow Isabelle Bousquette on LinkedIn, Instagram, X, and TikTok for more behind the scenes on her tech and AI coverage, and lately, her contributions to the WSJ Leadership Institute’s new Executive Resilience series, where she’s profiling America’s top execs about their fitness and wellness habits.

Follow Belle Lin on LinkedIn and X for her latest reporting on enterprise technology and AI.

Steven Rosenbush is chief of the enterprise technology bureau at the WSJ Leadership Institute. He also has a column. You can follow him on LinkedIn.

Tom Loftus is the editor of The Morning Download. He suggests following Isabelle, Belle and Steve on their various social channels. But if you insist, here’s his LinkedIn.