Slop, slop, slop. All I hear are complaints about people saying there is too much “AI slop.” All that code everywhere on GitHub, online, and bots everywhere—all that slop. I am not sure anyone has really stopped to think about the amount of slop we already have.
Us humans, we create code. We often leave it “as is” and push out a prototype of something that is usable but not the final product. Software has gotten to the point where it is never perfect and will always require updates. Broadband internet helped with that transition. Back in the day, before broadband, we had what us old people would call “Dial-Up” or a “BBS.” It would take a long time even to get text to show up on a screen, let alone push a software update. Because of that, you made sure your software did exactly what it was supposed to do from the moment it was installed throughout its entire use on the end-user’s computer.
Broadband was not just at home; we had it at work and in schools, too. While the enterprise had fast connections, it was only when broadband was introduced to people’s homes that “software updates” really entered the heads of programmers.
Mind you, we did have updates. For Windows, they were called “Service Packs” and usually came on a CD, or you performed a fresh install of Windows. If you had an update to grab online, it better have been really important because that download could take an hour. Interestingly, Windows can still take an hour to update today. As reported by Tech Republic, Microsoft released an update with 165 vulnerability fixes that included two zero-days. That is not good.
Some, including Steve Gibson from Security Now, believe Anthropic’s “Mythos” is the secret driver behind this sudden wave of updates. Gibson argues that —and that this partnership is exactly why we saw such a massive rollout of Windows updates this past week. He doesn’t think Anthropic is bluffing, either; he believes the tool is so powerful that they truly cannot risk releasing it to the general public just yet.
A 27-year-old bug in OpenBSD, the most used operating system on the planet for networking hardware. The point I am trying to make is that humans are not always good at making code. We are not perfect at what we do. Don’t get me wrong, humans are very good at making code—just look at how far we have gotten today and what we have built with what we have learned. However, we have now made something that can learn what we programmed these computers to do in seconds, and these things can produce really good results—but only if there are humans overlooking the code and the project.
The process of making what we need is also changing. We no longer necessarily need to know how to write Python or Bash scripts; we now just need to know how to be a good project manager. The reason you go to school now is to learn how to manage a team. If you can manage a team, imagine the possibilities. If you have ideas for a company or a project, you build a team. That team is now AI.
“AI” is a bit of a vague term because it covers everything from your phone’s autocorrect to Google Translate. A better way to think of it is as an AI Assistant that manages a team of Agents.
Think of an Agent like a specialized digital worker. Instead of you doing every little task, you act as the boss. You tell the Assistant what you need, and it delegates the work:
- Agent 1: “You write the code.”
- Agent 2: “You check that code for mistakes.”
It’s like having a group of experts on standby, each waiting for their specific marching orders.
Keep in mind that each agent is using an LLM model, which counts as usage depending on the system you use. If you use a provider like GitHub or Claude, your usage goes up. However, if you are hosting it yourself using LM Studio or AnythingLLM, the requests will simply queue up and process one after another.
I think we are in a transition toward something really big, and we are right in the middle of it. We are now in the middle of learning how we program and make software. Since the dawn of coding, it was always the same: code was everywhere and we really did not have any standards. Over time, we created and started using those standards. Now, we are learning the new standards in AI coding. Honestly, computers are better at talking to themselves and just making the machine do whatever it needs to get the job done.
Thank God we all do not have those dish-washing robots in the house that take 45 minutes to do anything, or else we would be up a creek! Well, I do not actually think it takes 45 minutes—I think I just heard that it takes a while. Anyways…
The companies that relied on software as their main source of revenue are in a lot of trouble because now we can all pretty much make our own. That software might not be perfect for everyone, but it may be perfect for you—and it may never even need an update.