Who is Aaron Levie?

Aaron Levie recommends six books as a complete framework for predicting technology outcomes: Seven Powers, Positioning, The Innovator's Dilemma, The Innovator's Solution, Blue Ocean Strategy, and Crossing the Chasm.

“If you if you had the time and you could be like locked in a room and read like six books, if you if you read that, you will be able to predict 100% of things that happen in technology like without fail.”

Aaron Levie explains why AI agents amplify rather than replace enterprise software: agents derive their power from accessing and traversing existing software systems and data sources, making those platforms more valuable as connective tissue.

“What why is Claude Tag so powerful? It's because it accesses your software systems that you can give it access to data. So box is one of those data sources as an example.”

AI agents are increasing usage of existing enterprise software such as Box rather than displacing it, because agents actively roam and access stored data.

“In our case, we actually see an increase in usage because agents are now roaming around accessing all of this data.”

Vibe coding will not replace existing enterprise software. It will layer on top of trusted data stacks, used by IT staff to customize workflows further.

“You look at Vibe coding and you say well Vibe coding must then replace the you know the software that we already use and probably the real answer is no it'll probably just be built on top of the software that we already use and so it'll be the IT person going and customizing their workflow even further but on a data stack that they trust is reliable and going to work you know very effectively.”

Aaron Levie on AI doomers and catastrophization.

“Most people most AI doomers should probably see a therapist because it's all it's all just catastrophization.”

Mandatory extensive review processes for AI model releases could kill iterative development cycles, pushing the industry toward larger and far less frequent updates.

“It'll mean, he writes that each model update will go through an extensive review, testing, and feedback process. And in that process, lots of groups will weigh in on the risk of the model and there will be lots of subjectivity on what the actual risks are or practicalities of exploiting those risks. He argues that if nothing else, this could significantly change the way we get model releases. Moving away from this paradigm of quick iterative releases to something that's much more irregular with bigger updates at one time.”

Aaron Levie on work expanding to fill AI productivity gains.

“The easier it is to do work, the more work you'll do and the more tired you'll be at the end of every day.”

AI will create more human work, not less. Human creativity and demand will generate new tasks to absorb every productivity gain.

“I just think we will find a way to create a ton more work for ourselves for better or worse.”

Jason Lemkin on why AI-title hires will fail at a high rate

“I think these hires, these Aaron leaving hires, are going to churn at a higher rate than CRO because they'll talk the talk, but they won't be able to deploy the agents.”

AI-title executive hires may churn faster than even CROs, the historically highest-churn executive role, because their inability to actually deploy agents will be exposed quickly.

“I think these hires, these Aaron leaving hires, are going to churn at a higher rate than CRO because they'll talk the talk, but they won't be able to deploy the agents.”