I agree with and second the main point you’re making about the current problem with major AI development. It doesn’t yet provide stability or meaningful cost reduction for most businesses. The constant updates and shifting capabilities create a real bottleneck for adoption, both in small companies and large enterprises, while costs are still hard to justify.
Another challenge is the sheer number of competing options. Most AI labs will likely converge toward similar capabilities over time, while open-source models may continue to hold an edge in areas like local privacy and decentralized control, which can be more attractive than relying entirely on cloud-based services.
Even in my own workflows the amount of time I spend to keep Claude Skills and MCP servers updated with the reality of my day-to-day is bigger than I'd like to admit. For knowledge work, time savings are more like 3-4x than the 10x some people report on fixed task with clearly defined contexts.
Perhaps on a collective level, but there are many great book authors I've followed for years whose recent postings clearly bear the traces of AI taking over their thinking (and writing) for them.
If it can happen to best-selling published authors, what does that mean for the rest of us who aspire to write something meaningful & enduring?
Skills atrophy and cognitive decline are definitely real, but also reversible. I'm actually looking forward to have the time to start writing novels again. One of the main pain points I had was the amount of time it took me to do research to get the settings right.
The thrill of discovery - even in a well-outlined story - of not knowing what your characters are going to do in the next paragraph won't change just because you've tasked agents with most of the research (although they might not find exactly the most salient details that would strike you yourself).
I think whether companies adopt agents as robot slot machines or not, dependence of a single company has to be avoided. I feel human in the loop will always lead to better outcomes, and that's why I agree organizations simply with access to the most advanced AI models won't win!
I agree with and second the main point you’re making about the current problem with major AI development. It doesn’t yet provide stability or meaningful cost reduction for most businesses. The constant updates and shifting capabilities create a real bottleneck for adoption, both in small companies and large enterprises, while costs are still hard to justify.
Another challenge is the sheer number of competing options. Most AI labs will likely converge toward similar capabilities over time, while open-source models may continue to hold an edge in areas like local privacy and decentralized control, which can be more attractive than relying entirely on cloud-based services.
Even in my own workflows the amount of time I spend to keep Claude Skills and MCP servers updated with the reality of my day-to-day is bigger than I'd like to admit. For knowledge work, time savings are more like 3-4x than the 10x some people report on fixed task with clearly defined contexts.
I'm not sure whether "knowledge" is irreversible.
Perhaps on a collective level, but there are many great book authors I've followed for years whose recent postings clearly bear the traces of AI taking over their thinking (and writing) for them.
If it can happen to best-selling published authors, what does that mean for the rest of us who aspire to write something meaningful & enduring?
How did we write before the advent of AI?
And what have we lost because of it?
Skills atrophy and cognitive decline are definitely real, but also reversible. I'm actually looking forward to have the time to start writing novels again. One of the main pain points I had was the amount of time it took me to do research to get the settings right.
The thrill of discovery - even in a well-outlined story - of not knowing what your characters are going to do in the next paragraph won't change just because you've tasked agents with most of the research (although they might not find exactly the most salient details that would strike you yourself).
There's also solid evidence that the cognitive decline caused by smartphone usage can be reversed in just two weeks: https://metacircuits.substack.com/p/the-attention-tax
Hopefully, teams will go on to take this knowledge and build their own versions.
Yes! Self-hosting is going to be a thing in 2026 (whereas in 2025 you'd have been laughed straight out of the room).
They’re still making memes about bosses asking their staff if they can build their own local agents.
That Douglas Adams quote really nails the psychology of how we handle shifting tech.
Right - it's a great insight!
The companies that win with AI won't be the ones who used it most. They'll be the ones who rebuilt around it.
Let's make sure for our clients becoming "AI native" has a different ending from building an "internet business" in 1999 :-p
I think whether companies adopt agents as robot slot machines or not, dependence of a single company has to be avoided. I feel human in the loop will always lead to better outcomes, and that's why I agree organizations simply with access to the most advanced AI models won't win!