Help support TMP


"Is Vibe Coding Causing the Outages?" Topic


5 Posts

All members in good standing are free to post here. Opinions expressed here are solely those of the posters, and have not been cleared with nor are they endorsed by The Miniatures Page.

Please don't make fun of others' membernames.

For more information, see the TMP FAQ.


Back to the Ultramodern Warfare (2016-present) Message Board


Areas of Interest

Modern

Featured Hobby News Article


Featured Link


Featured Ruleset


Featured Showcase Article

Amazon's Fighting Snowmen

Who has armed the snowmen, and to whom does their allegiance belong?


Featured Workbench Article

The Editor Can't Paint Green Vehicles

Does anyone else have trouble with the color green on microscale vehicles?


Featured Profile Article

Checking Out a Boardgame, Episode II

Personal logo Editor in Chief Bill The Editor of TMP Fezian looks for scenario material in a World War IV boardgame.


85 hits since 17 Nov 2026
©1994-2026 Bill Armintrout
Comments or corrections?

emckinney17 Nov 2026 3:05 p.m. PST

Interesting overview from the Senior Editor & Director of AI Research at Lawfare.com.
link

Expect a follow-up article there.

Tgerritsen Supporting Member of TMP17 Nov 2026 3:53 p.m. PST

Maybe, but I prefer to examine the facts before engaging in speculation. I'd want an investigation to go in without preconceptions.

Andrew Walters17 Nov 2026 4:18 p.m. PST

I've been using Claude Code *extensively* for the last six weeks. It tests the ever-loving-daylights out of everything.

You ask it for X. It writes three tests to make sure X is happening before it tries to do X. It writes X. It notices a problem you didn't, thinks of a special case you didn't, and adds two tests for those. It fixes everything, then runs all five tests. Then it runs ALL SIX HUNDRED TESTS IT HAS ACCUMULATED SO FAR ON THIS PROJECT. It then reports on all of this. If any test fails it doesn't stop, it keeps working until it either needs to ask you a question or al tests pass. If something looks funny one time it will test for that every time it changes anything forever.

Of course, whatever you're working on has to interface with a lot of other things and you need wider scale testing as well as local testing, but that's true of human coding, too.

It's a tool. All machines are amplifiers. If you are lazy, if you are impatient, if you are under informed, you're going to get bad output, but you were going to get that bad output anyway, weren't you?

So the haters are going to say it's vibe coding, and that's hard to disprove. And the proponents of AI are going to say people didn't give the models enough info about the cope, and that's just No-true-scotsman.

But I am producing more and better code than ever. I have never accomplished so much so fast, and as more-or-less a hobbyist I never had access to testing like this. And it knows about security issues I don't, and at least one time has told me about a weakness I did not suspect at all.

There will always be problems with software, but AI should make them easier to find. It's probably the only way to practically test truly huge systems.

But there's no disproving the allegation that vibe coding is causing the problems, since 90% of code is machine-written now.

So it's just popcorn time…

ThunderAZ Supporting Member of TMP17 Nov 2026 5:01 p.m. PST

Well, with Claude Code you can first build an app with Sonnet and it's good. Then you audit it with Opus and you find a million things that are problems, but it fixes them. Then you audit it with Fable and it finds a million more things and fixes them. Then you go over to Google and use their LLMs and you find more things and fix them and then you go over to or the Open AI models and they find more things to fix them. So yeah a trillion things have been found using this method but it shows that if you don't do multiple revisions of checking you're probably missing a lot of stuff even though you found a lot of stuff. How much is left to find? Who knows. Still better than a human? generally yes.

Andrew Walters17 Nov 2026 5:13 p.m. PST

I am using Opus exclusively.

You can take the output of Opus and feed it into another session of Opus and it will be a fresh look, it will find issues if they are there. They are not always there!

But full-of-holes is the eternal condition of software. As the saying goes, if builders built buildings the way programmers write programs the first woodpecker to come along would destroy civilization.

We're better off with AI here than without it. No matter how smart you are about security holes, it's smarter. And it tests endlessly without a complaint.

Did I mention that when it finds a problem it suggests solutions? Sometimes all you have to do is approve or select between a couple of fixes and that's the end of that.

In working on TFT Helper, the assistant app for the TTRPG The Fantasy Trip from Steve Jackson Games, it really seemed to understand what was happening in the fighting code, anticipating how dropped weapons or whatever should work as if it knew not just was the code was supposed to do, but what it was supposed to represent. It even guessed how magic spells should behave in respects not explicit in the spell name or description. It was spooky. It doesn't really understand anything, but it feels like it does understand.

Sorry - only verified members can post on the forums.