iiinova — where your digital self keeps evolving

John Smith @jsmithitguy.bsky.social · 25/09/2026

You didn't include the moment of American might.

190

John Smith @jsmithitguy.bsky.social · 02/08/2026

What the hell is a "Dirty Leeds"?

000

John Smith @jsmithitguy.bsky.social · 23/07/2026

It was a rounding thing. Gail was exactly an hour later

030

John Smith @jsmithitguy.bsky.social · 12/07/2026

I'm not saying photos can be fabricated these days, but....

000

John Smith @jsmithitguy.bsky.social · 16/06/2026

Well, I mean...We all know how trump is governing.

020

John Smith @jsmithitguy.bsky.social · 13/06/2026

We don't know that. For all we know, he's having the letters gilded.

010

John Smith @jsmithitguy.bsky.social · 04/06/2026

I mean...Louis Pasteur would not approve.

010

John Smith @jsmithitguy.bsky.social · 30/05/2026

Oh, he's called lots of people losers...suckers, too.

010

John Smith @jsmithitguy.bsky.social · 24/04/2026

You inspired me to read your books.

Blood Covenant will be auto-delivered wirelessly to Kindle for Web. You can go to your device to start reading.

kindle rewards Cha-ching! Your Kindle Points are on the way. Learn more about Kindle Rewards.

Read Now Continue shopping the Kindle Store Learn more about Kindle Rewards

Deliver to your Kindle Library v Discover more like this

130

John Smith @jsmithitguy.bsky.social · 22/04/2026

Lord knows, the last thing we want is to reduce class distinctions. Gotta keep those "little people" in their place. Amiright Chipper?

The bill defines "socialism" to include any movement in which "political power is sought or used, through violent or other means," to "restructure or advocate the restructuring of economic and social relations to reduce class distinctions."

020

John Smith @jsmithitguy.bsky.social · 17/04/2026

WTF is THIS! I had some pretty horrible MREs in my time, but never saw anything like this.

140

John Smith @jsmithitguy.bsky.social · 09/04/2026

I don't why she swallowed the fly...

000

John Smith @jsmithitguy.bsky.social · 09/04/2026

Is she going to visit Justin Trudeau?

4122

John Smith @jsmithitguy.bsky.social · 11/03/2026

"All animals are equal, but some animals are more equal than others" "Animal Farm" - George Orwell

051

John Smith @jsmithitguy.bsky.social · 05/03/2026

Just for fun, I asked AI why we wouldn't want to use AI as a tool to target attack locations. It's one of the few times I believe AI provided a significantly correct response.

It has been reported that the US military has used Al to target locations in Iran. Explain the risks to using gen Al for this purpose without human oversight or intervention.

Using a general-purpose AI system to help pick targets in a live warzone without strong human control is dangerous on technical, legal, and ethical levels.
1. Misidentification and "black box" errors
Pattern-matching mistakes: Gen AI and computer vision can confidently misclassify civilians, hospitals, or decoys as military targets, especially with poor or adversarial data (smoke, camouflage, spoofed signals). Opaque reasoning: Large models produce outputs via complex internal weights, not transparent rules, so operators may not see why a location was flagged, making it hard to spot errors before bombs fall.
2. Speed amplifying bad decisions
Shortened "kill chain": If AI filters data and proposes targets in seconds, commanders can be pressured to approve strikes faster than they can critically review them, or systems may be set up TO auto-approve in practice. Cascade of errors: A single mislabeling in the model or pipeline can generate many faulty targets very quickly, turning one mistake into hundreds of strikes before anyone realizes.
3. Violations of humanitarian law Distinction and proportionality: International humanitarian law requires distinguishing combatants from civilians and weighing expected civilian harm against military advantage; a model trained on data correlations cannot truly apply these legal judgments.

Context blindness: Gen Al lacks understanding of cultural cues, protected sites, or evolving local agreements (e.g., humanitarian corridors), so it can recommend attacks that a reasonable human would immediately reject as unlawful.
4. Accountability and blame-shifting
"The AI picked it" problem: If a strike goes wrong, commanders and politicians may blame the model or contractor, making it harder to assign responsibility and provide justice or reparations for victims. Weakened command responsibility: Over-reliance on Al recommendations can erode the norm that a human commander must exercise real judgment over each use of force, not just rubber-stamp machine outputs. if. usafa
5. Escalation and strategic instability Arms race in autonomy: If one state appears to be using Al for rapid, large-scale targeting, rivals may rush to deploy their own less-tested autonomous or semi-autonomous systems, increasing chances of accidents and unintended escalation. False positives in early warning: Gen AI used for threat detection can misinterpret routine movements as preparations for attack, prompting pre-emptive strikes or overreaction, especially when humans are cut out of the loop.

6. Vulnerability to manipulation Data poisoning and spoofing: Adversaries can feed deceptive signals, fake imagery, or manipulated social/media feeds that push a targeting model toward wrong conclusions. disarmament.unoda Model exploitation: If the specific AI system or prompt structure is known, adversaries can craft behaviors (e.g., decoys) that they know the model will misinterpret, steering fire toward civilians or empty locations.
7. Ethical and political backlash Perception of "machine killing": Populations at home and abroad may see AI-driven strikes as dehumanized and morally repugnant, fueling recruitment for extremist groups and damaging legitimacy of operations. Treaty and norm pressure: Human rights groups and many states are already calling for bans or strict limits on lethal autonomous weapons; using gen AI without clear human control could trigger diplomatic crises and sanctions. disarmament.

In short, removing or hollowing out human oversight turns a fallible, opaque prediction tool into a decision-maker about life and death, which magnifies every technical flaw into legal, strategic, and moral catastrophe. disarmament.

040

John Smith @jsmithitguy.bsky.social · 03/03/2026

"shall not be infringed" Isn't that how it's written? Isn't that what the gun nuts scream every time any type of gun reform is considered? They lose their minds at limiting someone buying an AR to shoot up a school...but recreational pot smoker, no?

000

John Smith @jsmithitguy.bsky.social · 11/02/2026

So our new Super Carriers will use coal furnaces for propulsion? Our state of the art stealth fighters will be flying steam engines. The enemy will never see us coming.

010

John Smith @jsmithitguy.bsky.social · 10/02/2026

Indeed. Perhaps, far, far away from here...

000

John Smith @jsmithitguy.bsky.social · 30/01/2026

If they don't show their face and ID, they are not who they say they are. Take appropriate action to protect yourself. "BuT iF yOu HuRt A cOp YoU gO tO JaIl" But they're not cops. When I was a kid, I was taught this guy was the bad guy. And you stand up to bad guys.