Know what's running.
3 SIGNALS · 1 LESSON · 1 ACTION · ABOUT 5 MINUTES
AI tools and alarming headlines can create the same bad habit: acting before thinking.
This week, the useful move is to slow down long enough to identify the problem, review what you already have, and prepare for what might actually happen.
SIGNAL 01
Start AI adoption with the problem, not the product
What happened. Businesses are being offered a growing menu of AI tools, but tool selection is not the first decision. The first decision is which costly, repetitive, slow, or error-prone problem is worth solving. Only then should you review the software you already pay for. For example, if you’re in the Microsoft or Google ecosystem, check what your subscription already includes: eligible Microsoft 365 plans come with Copilot Chat, while Google has added Gemini features to many Workspace editions.
Why it matters. Buying a separate tool before checking your current stack can add cost, duplicate capabilities, scatter company data, and give employees one more system to learn. Embedded AI will not always be the right answer, but it is usually the right place to begin the comparison.
What to do. Write down one specific workflow problem, its current cost, and the result you want. Then ask your existing provider what your license includes, what data the AI can access, and which administrative controls are available before you shop for anything new.
SIGNAL 02
California is preparing for El Niño before the storms arrive
What happened. NOAA says El Niño is strengthening, with a greater than 90% chance of a very strong event through fall and winter 2026–27. On September 21, California proclaimed a statewide emergency to speed preparations for possible flooding, landslides, debris flows, coastal impacts, road closures, and power outages. El Niño raises the odds of a wet Southern California winter, but it does not guarantee one.
Why it matters. The uncertainty is a reason to prepare, not a reason to wait. Organizations should know which services must continue, who can make decisions, how employees and customers will receive updates, and which suppliers or systems could become unavailable. Service providers should be ready to explain what help they can deliver, under what conditions, and in what order when many clients need assistance at once.
What to do. Review your continuity and severe-weather plans now, then conduct a tabletop exercise. FEMA offers a winter-storm exercise and continuity exercise starter kits. Test communications, remote work, backup power, critical vendors, recovery priorities, and decision authority. If you serve other organizations, prepare your staffing, contact lists, response packages, and client instructions before the forecast turns into an incident.
SIGNAL 03
Cal Newport is helping separate AI risk from AI mythology
What happened. On the Better Offline podcast, Ed Zitron spoke with computer scientist Cal Newport and journalist and astrophysicist Adam Becker about the ideas and communities behind modern AI doomerism. The episode, “Stopping The AI Safety Cult”, is deliberately provocative, but the useful question beneath the title is serious: Which AI claims are supported by current technical evidence, and which depend on speculative stories about future superintelligence?
Why it matters. Newport has been a steady voice of reason through the AI hype and doom cycle. He does not argue that AI creates no danger. He asks us to distinguish demonstrated harms, weak controls, and reckless experiments from claims that today's large language models are inevitably becoming autonomous superintelligence. That distinction leads to better risk decisions.
What to do. Listen to the episode, then read Newport's case against “doom trolling”, his analysis of reports that AI agents had “gone rogue”, and his argument that chasing superintelligence can cause harm even if superintelligence remains a fairy tale. For each alarming claim, ask what happened, what evidence supports the interpretation, and what concrete control would reduce the demonstrated risk.
THE LESSON
Good decisions begin before urgency takes over
The common failure in all three stories is letting the proposed solution, the headline, or the prediction choose the problem for us.
An AI vendor shows a polished demo, so the organization starts searching for a use. A storm warning appears, so teams discover that their continuity plan contains old phone numbers and an owner who left two years ago. An AI executive describes an extinction scenario, so practical questions about access, testing, monitoring, and accountability get pushed aside by a debate about science fiction.
Clear thinking starts in the opposite direction. Define the outcome. Inspect current conditions. Separate what is known from what is possible. Identify who owns the decision. Then test the plan while the consequences are still inexpensive.
Preparation is not panic. It is how organizations preserve room to make good decisions after the pressure arrives.
DO THIS WEEK
Run a one-hour “problem, plans, and premises” review
Bring together the people who own operations, technology, security, and communications. Spend 20 minutes on each question:
Problem: Which workflow problem are we trying to solve with AI, and what capability have we already purchased?
Plans: Which essential service is most exposed to severe weather, and when did we last exercise its continuity plan?
Premises: Which alarming AI claim is influencing our decisions, and what evidence and assumptions sit underneath it?
End with one owner and one next action for each answer. A meeting without ownership is just group note-taking.
Hackademic Solutions can help your team define an AI use case, review the capabilities and controls in your current technology stack, or conduct a severe-weather and business-continuity tabletop. The goal is a practical decision and a tested plan, not another binder for the shelf.
— Chris Simpson
Hackademic Solutions
The Hackademic Briefing · Know what's running.