What would Wiener think of the current human use of human beings? He would be amazed by the power of computers and the internet. He would be happy that the early neural nets in which he played a role have spawned powerful deep-learning systems that exhibit the perceptual ability he demanded of them—although he might not be impressed that one of the most prominent examples of such computerized Gestalt is the ability to recognize photos of kittens on the World Wide Web.
Compression made the information age possible by stripping things down to fit the pipes. Expansion made the AI age possible by blowing data back up again. Both operations leave marks; we’ve learned to spot compression artifacts, but we’ve only just begun to reckon with expansion artifacts. Until we do, there’s a lot of risk to manage.
There’s a fundamental problem with these tools beyond the capacity of any deployment strategy to solve: the tool requires expertise to validate, but its use diminishes expertise and stunts its growth. How does one become an expert? There are no shortcuts; there is only continuous hard work and dedication. I was once told of writing, great writers learn how to break the rules in new and ingenious ways by first learning the rules.
But how is a new developer meant to learn the rules if their day-to-day work is nothing but the babysitting of models? How will they gain the hard-won experience that allows a human in the loop to be a useful safeguard?
These models alter cognition in ways deleterious to human prosperity. In other words, for as much output as they provide, they take something important from us.
I can’t remember the last time a blog post resonated with me this much.
Craig’s criteria on his job search:
- One: fuck offices
- Two: fuck AI
- Three: fuck React
And his conclusion:
Fuck work
Eleven years ago, I wrote:
Sometimes I consider the explosive growth of computation and think that strong AI is a near-term inevitability.
Then I remember printers.
That was just a brainfart, but Robin tackles it seriously in his thoughtful essay.
A pleasing image: if indeed AI automation does not flood fill the physical world, it will be because the humble paper jam stood in its way.
Software cannot, in fact, eat this world. Software can reflect it; encroach upon it; more than anything, distract us from it. But the real physical world is indigestible.
Related to Matt’s thoughts:
…working with agents feels much less like classic deep work, and much more like playing a game. Not to say the work is frivolous—it’s just because it feels like I’m in a game loop.
Flow, at least in the usual sense for me, feels smooth and continuous. The work and your attention starts to line up so cleanly that the experience becomes frictionless. You disappear into the work and meld with it. One notable aspect of flow has been I lose track of time. Working with agents on the other hand, is not like that at all. It’s highly engaging, but in a more jagged, reactive way. I’m focused, but not settled. I’m absorbed, but not merged with the task. I’m paying close attention the whole time, but the attention is dynamic and tactical rather than continuous. I don’t lose track of time at all.
Compression, transformation, expansion.
Knock, knock! Who’s there? Control freak (now you say “control freak who?”)
Brian Eno on prototyping and fidelity.
Wake me up when we get to the plateau of productivity.
A large language model is as neutral as an AK-47.