Retrying an automation that failed earlier this year with Grok Bot, this time it worked first try
The author spent many tokens early this year trying to automate a task with AI but it was never reliable, then retried it with Grok Bot and found it effortless, arguing that AI improves so fast that hard tasks are worth waiting on.
The "Far Centaurus*" problem of interstellar travel says that you should wait to launch a trip to another star system because rocket tech will advance and later rockets will overtake earlier rockets. I just experienced that with AI, which is improving so quickly. I spent a bunch of time and tokens at the beginning of this year trying to use AI to automate something, but it was buggy and never worked reliably. Eventually I dropped it. I just tried it again with Grok Bot and it was effortless. If you procrastinate with AI, you may be rewarded. If something is too hard right now, just wait a little while. *see Andrew Kennedy’s 2006 paper in the Journal of the British Interplanetary Society, “Interstellar Travel: The Wait Calculation and the Incentive Trap of Progress.”