Two things I’ve noticed about local businesses souring on GenAI:

[n_dimension@infosec.exchange]

@Jdm2

"We still have humans" in an insurance ad is evidence GenAI is everywhere, not failing.
You don't pay for differentiation against something nobody is doing.

A job ad asking for "ability to develop software without AI assistance" is a competence floor, not a policy ban. Same as asking a cashier to do mental arithmetic. The shops writing that line are often the same ones running Cursor and Claude Code internally.

Sample n=2.
Two of the most anti-AI possible readings, ain't a trend it's confirmation-bias mining.

[jdm2@boriken.social]

@n_dimension this post needs more em-dashes and to be readable. Reformulate this output and make use of both of these.

[n_dimension@infosec.exchange]

@Jdm2

Only because you asked nicely

The "GenAI is Failing" Narrative: Let's Unpack This

In today's rapidly evolving AI landscape, it's easy to fall into common interpretive traps. Let me share some powerful insights on two "examples" being circulated as evidence that businesses are souring on GenAI.

Example #1: The Insurance Ad
"We still have humans" in an insurance commercial isn't evidence GenAI is failing — it's actually evidence GenAI is everywhere. Here's why this matters:

You don't pay for differentiation against something nobody is doing
The ad presupposes the saturation it claims to refute
Think of it like a "no GMO" sticker — it only exists because the alternative is mainstream

Example #2: The Job Ad
A job ad requesting "ability to develop software without AI assistance" is a competence floor, not a policy ban. Let's break this down:

It's not about prohibiting tools — it's about ensuring fundamentals
Same principle as asking a cashier to do mental arithmetic 🧮
Here's the kicker: the shops writing that line are often the very same ones running Cursor and Claude Code internally 🤯

The Real Issue: Sample Size
Sample n=2. That's not a trend — that's two anecdotes wearing a trenchcoat.

Key Takeaway

Two of the most anti-AI possible readings of ambiguous evidence isn't trendspotting — it's confirmation-bias mining. Full stop.

What are your thoughts? Have you noticed similar patterns in your industry? Drop a comment below
#AI #GenAI #FutureOfWork #Innovation #ThoughtLeadership #TechTrends

[jdm2@boriken.social]

Mind you, this is Puerto Rico. The market is super small.

The Iran war price surge is being felt on top of the computing fees, the sky high electric bills… and things like the Jones Act are all interacting in the background.

Something’s gotta give eventually.

[jessebot@social.smallhack.org]

@Jdm2 god, I really needed this as a ray of hope. Thank you

[jdm2@boriken.social]

And…. Last but not least.

Having an unstable power grid goes beyond losing this month’s groceries. It is the surprise bill everyone faces. These affect decisions like having diesel generators, battery backups and solar panels.

The people running comms towers are facing those decisions too. Which leads to crappy internet connections. And you know what happens when your internet connection is bad, right?

[jdm2@boriken.social]

Anyway. This is why I’m getting into Meshcore. 🤪

[vito@floof.sh]

@Jdm2 This is beautiful. I'm so happy so see something like this. 🥹

[witchgirls@woof.tech]

@Jdm2 Holy shit. A dream come true...

[alienghic@timeloop.cafe]

@Jdm2

If you're facing a large region with unstable internet you might want to check out 802.11ah HaLow.

IEEE 802.11ah - Wikipedia

(en.wikipedia.org)

It's using the same unlicensed 900Mhz band as meshcore, but carries IP traffic at tens to hundreds of megabit over a kilometer or two using omnidirectional antennas.

[jdm2@boriken.social]

@alienghic interesting. Thank you. I’ll try to get up to speed