The greater good

Matt Shumer gives his take on where we are with AI in Something Big Is Happening. This article has been viewed 67 million times (as of right now) and he’s being pummelled by the haters on one side calling it AI slop, and applauded by the other half. The counter argument appears to be an essay called Tool Shaped Objects, which of course half the audience are celebrating while the other half are calling it slop.

In our polarised new world, what did you expect?

Slop or not, I think Shumer’s point (AI is replacing service-working humans) is probably right. Less than a year ago people thought it would take years, if ever, for software engineers to be made redundant by AI. From what I can see (using these tools to “write” code most days) it’s pretty much done. We’re LLM supervisors now, we don't write code. Until supervision is no longer required. 

For those embracing it – either devs at forward-thinking companies or indie builders creating their own product – it’s an incredible time to be alive. The creative possibilities are enormous, and the increase in velocity is real. If you have the time, permission and inclination you could be riding the crest of the wave. You can build things you never dreamt possible. Cathedrals in the sky. Such is the rapid advancement, your mind will be blown daily. 

That’s my experience, at least.

Coding is the testing ground. It’s at the heart of AI companies, it’s what they know best, so they set out to solve that first. Now they‘ve accomplished that, it makes sense for them to focus on other sectors.

Shumer (or his LLM 🤷) notes:

Law, finance, medicine, accounting, consulting, writing, design, analysis, customer service. Not in ten years. The people building these systems say one to five years. Some say less. And given what I've seen in just the last couple of months, I think "less" is more likely.

Most people don’t yet understand this because they’ve only used the free ChatGPT chatbot as a replacement for Google, and not experienced the capability of agentic AI. There’s a seismic shock coming as the new normal of AI automation arrives to other industries. And it’s hurtling towards us. It seems naive to suggest otherwise.

Nobody knows what will actually happen, but a sea change is inevitable. If you haven’t used them in anger yet (and budget permitting), I would recommend spending $20 to give the latest models a go (Claude Opus 4.6 or GPT-5.3 Codex). Throw complex work problems at them, learn what’s possible, and prepare to have your mind blown. You’ll think differently, quickly, I’m sure. If our jobs are going to be transformed by these tools, it’s probably worth becoming adept at using them sooner rather than later, right?

Knowledge is power, as it has always been. In this new world we're no longer the most knowledgeable in the room, but we've been handed an immense new power that we need to get used to. We just need to make sure we use it for the greater good.