Last week I wrote about the frenetic pace of progress in AI and the realization that we live in a world where answers to most questions are at hand but are we asking the right questions. In a world of LLMs, the "answer" is a commodity. The "question" is the intellectual property. If answers are there for the asking, clearly the skill that is needed is a deep ability to ask the right questions.

So leaders need to ask better questions - they need to get better at the art of asking questions. In the article “The Surprising Power of Questions¹”, authors Allison Wood Brooks and Leslie K. John taught us that questioning is a powerful tool for unlocking value in organizations. They say “Much of an executive’s workday is spent asking others for information—requesting status updates from a team leader, for example, or questioning a counterpart in a tense negotiation. Yet unlike professionals such as litigators, journalists, and doctors, who are taught how to ask questions as an essential part of their training, few executives think of questioning as a skill that can be honed—or consider how their own answers to questions could make conversations more productive”.

This second article in HBR² provides a framework targeted towards strategic decision-making in complex environments. Briefly:

It provides a framework with five domains (investigative, speculative, productive, interpretive, and subjective) - by covering all of them, a leader avoids blind spots (our expertise often blinds us to aspects of a problem in the heat of the moment).

Generate your own questions for each category (or adopt those from the article) and for your next high stakes decision, ensure you have no blind spots by systematically asking questions from each domain. Every leader has a question mix they are comfortable with. If you only ask Investigative questions, your team becomes a "fact-finding" machine but loses its creative edge. If you only ask Speculative questions, you may suffer from "visionary fatigue" without ever reaching the finish line. If you are aware of your own blind spots, and know you are uncomfortable asking some questions, practice them in low-stakes environments.

We are in an era where everybody has access to all the answers and the winners will be the ones who can ask the right questions.

Google announced the release of Gemini 3.1 Pro with a 1M token input context window and a 65k output token limit representing substantial increases in memory and efficiency. Personally this made me want to go back to some old coding projects and am really having fun with them (now I just have to watch the agent manager plan, review and make some edits and viola):

This was primarily to see how to build using Antigravity’s agent manager and see how Claude Sonnet 4.6 and Gemini 3.1 performed. I am afraid, now I have to go back and review the code to see if I am even capable of evaluating the code quality differences. I have a better personal understanding of the power of these tools and am now excited about building more.

This week I was reminded that Volt Typhoon’s pre-positioning and the systematic preparation for kinetic-grade disruption continues. Although the U.S. government conducted a operation in December 2023 to dismantle the KV Botnet that didn’t seem to deter them - they are pretty resilient. Recent reports indicate that the botnet has been revived and is again being leveraged by the threat group. This resilience underscores the ongoing vulnerability of the "network edge," where insecure devices continue to serve as a low-cost, high-reward launching pad for state-sponsored attacks.

This news below is the one that caught my eye this week and the fact that US is not alone in this fight.