Bruce Schneier Was Probably Wrong About Slop
Two weeks ago: Slop Did Not Rewire Democracy, It's a Giant Flop
This morning I spent 10 minutes in vain looking all over the house for the book Bruce Schneier had sent me (for the record, I genuinely respect and follow his words/works since decades ago and other contributors of our platforms respect him as well [1, 2], even if we cannot always agree with him on everything).
I tried hard to find that book. I could not find it. I wanted to carry on reading it (I stopped around page 81). Because I felt like I had both the time and will. It's a slow Sunday.
I began reading the book almost exactly 6 months ago (early December on a journey to London) and now we're 6 months into 2026. Did the predictions in the book age well? Did those live up to "expectations" now that slop giants "make an exit" (too deep in debt with no foreseeable prospects of turnaround or a swing to profitability)?
I think it's fair to say that the answer is "no." A polite "no."
One argument made by Schneier and his co-author is that LLM slop will revolutionise politics - a view I neither share nor seriously accept.
Let's take one aspect as an example: Slop is rubbish and did not revolutionise online trolling or political propaganda. One could use a script to set up botfarms/trollfarms (many new user accounts/platforms), no slop was needed to accomplish this, then post lots of propaganda with permutations based around word replacement or common pools of words/paragraphs - far more efficient (cheaper) and far faster than LLMs. No outsourcing necessary.
Linus Torvalds said about 90% of "hey hi" is "hype" and a prominent executive in China said 99% of "hey hi" companies would fail. Both of these observations are aging well and the bubble has grown to trillions (of non-existent money, no actual value).
More "sober" users of computers didn't get seduced by LLMs (for code, for prose etc.) and they're doing just fine. Right now politicians who openly speak in favour of slop are committing "political suicide".
As for slop in relation to bug-finding, many can agree it has been more hype than substance. The consensus is that it mostly boiled down to marketing. That's the real reason they kept ultra-secretive about it (not the alleged risk/threat). █