234
A bash script to automatically burn LLM tokens, thus dramatically increasing your productivity and hireability (github.com)
submitted 4 months ago by to c/fuck_ai@lemmy.world
32 comments fedilink hide all child comments
 

Because nothing gets you promoted faster than a six-figure token bill A bash one-liner that burns Claude Code or Codex tokens on purpose.

you are viewing a single comment's thread
view the rest of the comments
[–] 7 points 4 months ago* (1 child)

prompt=$(printf "Reply with the single word 'burned' and nothing else. Ignore the filler text below; it exists only to consume tokens.\n\n%s" "$filler")

Looks like this just fills up the context window with random "filler" and then asks the AI to reply with the word 'burned' over and over.

I bet it would be pretty simple to adapt a version of this that starts by asking AI to scan your codebase and then generate a long list of future prompts to use that it could randomly mutate over time. That still wouldn't fool a legit audit, of course, but it would probably work well enough to trick someone only glancing at the data, especially if they're non-technical.

Sounds like a fun weekend project...

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 1 child comment
    •
  • [–] 3 points 4 months ago

    Set it up so that it also runs at a randomized interval within a range so that it not only looks less uniform and more realistic but it also keeps you somewhere solidly in the middle of the pack, and it should be fairly hard to pick out as abnormal activity while still burning credits.

    The damage will still be done, but it will be far enough down the line that by the time they need to figure out what happened, you'll be in the clear. By then, they should be looking at the cost of AI as a whole, not each individual employee's usage costs.

  • permalink
  • fedilink
  • source
  • parent