THE LOGICAL BOX
AI news & training for business owners & operators. One email. One clear next step.
THIS WEEK IN AI
A few months ago, Meta paid two billion dollars for an AI company. This week it started tearing that deal apart, because a regulator on the other side of the world told it to. The tools you lean on every day are not as permanent as they look, and that's worth a few minutes of your time.
In this issue:
Beijing forced Meta to unwind its $2 billion Manus deal
Anthropic asked the whole industry to plan for a pause
OpenAI filed to go public, chasing a trillion-dollar valuation
The Deep Cut: what breaks if your one AI tool disappears tomorrow
THE SIGNAL
What happened in AI this week
Image Source: OpenAI by Andrew Keener
What happened:
Meta bought Manus, a Chinese-founded AI startup, for about two billion dollars in December. Manus had moved its headquarters to Singapore first. In April, Chinese regulators ordered the deal reversed, and this week Meta finished the split. It cut Manus off from its internal systems, told staff to stop using Manus tools, and started moving projects onto its own platform. The founders are now trying to raise around a billion dollars to buy the company back.
Why it matters to your business:
If a tool's roots trace back to China, it can carry a shutdown risk that has nothing to do with how good the tool is. Any AI you build a daily process around should have a backup you can move to without losing a week of work.
Source: CNBC
Image Source: OpenAI by Andrew Keener
What happened:
On June 4, Anthropic published a proposal called "When AI builds itself." It argues that AI systems are getting close to improving themselves faster than people can keep up, and it wants the major labs to agree on a way to slow down or pause frontier development if safety falls behind. Anthropic said more than 80 percent of the code in its own systems is now written by its AI, not by its engineers.
Why it matters to your business:
When the company building the tool says the pace worries them, treat that as a signal about stability, not just safety. Plan your AI use around what works today, not around a roadmap that could shift under you.
Source: The Washington Post (AP)
Image Source: OpenAI by Andrew Keener
What happened:
On June 8, OpenAI confirmed it confidentially filed for a US IPO, a week after Anthropic did the same. Reporting puts OpenAI's target near a trillion dollars. No timeline yet, and the company said it may wait.
Why it matters to your business:
The companies behind your AI tools are becoming public companies answerable to shareholders. That usually means pressure on pricing and a push toward higher-margin plans. If your budget assumes today's prices hold, leave room for them to climb.
Source: CNBC
THE DEEP CUT
What it actually means for your business
Here's the question I keep coming back to after this week.
If the one AI tool your team leans on disappeared tomorrow, what breaks first?
For a lot of small businesses, the honest answer is more than they'd like.
Last year I watched a contractor run his whole estimating process through a single AI tool. It took his estimates from three hours down to about twenty minutes. Real savings, and he was right to be happy about it. But when I asked what he'd do if that tool changed its pricing or shut off access, he didn't have an answer. The entire gain lived inside one app he didn't control.
That's the risk this week put on display. Meta paid two billion dollars and still couldn't keep Manus. A regulator made the call, and the businesses using Manus for daily work woke up needing a new plan.
Most small businesses aren't out buying AI companies. But the lesson holds at every size. When a workflow depends entirely on one outside tool, you've handed control of that workflow to someone else.
This is where the operations side matters more than the tool side. The fix isn't to avoid AI, and it isn't to bolt on three backup tools you'll never open. The fix is to know where your single points of failure are, and to make sure the work itself can survive a switch.
In my day job doing systems analysis, the first question we ask about any tool is what happens when it goes down. Not if. When. That habit transfers straight to AI.
So before you go deeper, get clear on one thing. Which steps in your operation now run through a single AI tool with no fallback? If that tool vanished, would you lose an afternoon or a month? You don't have to fix all of it this week. You just need to know where you stand.
THE MOVE
One thing you can do this week
Run a quick AI dependency check on your own operation. About 15 minutes.
List the tasks your team now runs through an AI tool. Be specific. Drafting proposals, summarizing calls, writing customer replies, building estimates, whatever it is.
Next to each one, write the tool it depends on.
Mark any task where that tool is the only way the work gets done right now. Those are your single points of failure.
For the top one or two, write a one-line fallback. What would your team do next week if that tool was gone? Even "go back to the old template" counts. The point is to have an answer before you need one.
You are not trying to remove AI. You are making sure the work holds if a tool doesn't. Forward this to whoever owns your tools and have them fill it in.
THAT’S A WRAP!
If your team is leaning on AI tools and you're not sure where the gaps are, that's the first thing I look for in a free AI Savings Scan. Twenty minutes, no pitch. You walk away with a one-page snapshot: your biggest time drain, where AI actually fits, and the next move to make.
Thanks for reading,
Andrew Keener
AI Advisory, Training & Speaking for Leaders and Their Teams
Keen Alliance Consulting
Please share The Logical Box link if you know anyone else who would enjoy!
