Isn‘t this an example for something that will pay off later? Public baths come to mind. Those generally don‘t make sense from an economic point of view and are prohibitively expensive for private owners to build and maintain.
But they pay off by keeping society clean and healthy and prevent loss of workforce from mental or physical disease?
> In many cases the skills are available in house to do the necessary vetting, but these people are already overwhelmed with their existing day to day.
This is an interesting topic. We treat vetting output the same as doing the work ourselves, but that is not the case.
Doing the work is not the same as reviewing work done by others.
I have heard reports of software engineering companies that have gone full agentic. Their seniors only review stuff written by LLMs and it burns them out, because they have to switch context constantly.
I find this interesting because part of being a senior developer is that you are experienced enough that you won‘t make grave mistakes anymore. This is the case in many professions: you are relied upon to not make grave mistakes.
But those same people are now swamped with stuff that they are not able to review, so they will let a grave mistake slip through at some point.
> but I am absolutely in the camp of "there are many tasks where pattern matching is just as effective as actual understanding“
What if „being effective at something with pattern matching but not understanding it“ just means that you have identified only 90% of patterns and keep failing to learn the rest for whatever reason.
I was originally sceptical of LLMs and am far from the „agents will magically fix our future“-crowd, but sentences like these trip me up:
> „But pattern‑matching is not system understanding, and plausibility is not correctness.“
Why not? Who says that? Who proved that system understanding is not just more complex pattern matching?
> „LLMs predict tokens, not consequences“
Same here. LLMs output tokens but who says that they don’t form some internal group of token-predicting tensors that move together and constitute the internal model of a „consequence“? It is like saying humans don’t have thoughts, they just have electrical impulses moving their tongues.
I too think that LLMs seem to be a very specific form of intelligence, maybe resembling the parts of our brain that do language-processing, but it is a fact that they at least fake intelligence very convincingly. And that we actually don’t know how they do it.
It's true that there are few obvious bugs, but there are many subtle bugs now, usually outside the main interaction flows. 5 years ago, apple's software felt bugfree, essentially.
examples:
- if I change a note on my iphone and wake up my mac, I need to restart the notes-app before it syncs the change.
- if somebody leaves me a FaceTime video-message, I get an "unread"-badge that doesn't get away after I watch the video. There are multiple ways to get to that video and only one of them clears the "unread" badge.
- if I add a pronounciation field to a contact in my iphone, SIRI stops working and I need to restart my iphone to get it back.
A similarly scoped book series is „AI game programming wisdom“, which contains a multitude of chapters that focus on diverse, individual algorithms that can be practically used in games for a variety of usecases.
With databases there exists a clear boundary, the query planner, which accepts well defined input: the SQL-grammar that separates data (fields, literals) from control (keywords).
There is no such boundary within an LLM.
There might even be, since LLMs seem to form adhoc-programs, but we have no way of proving or seeing it.
The principal security problem of LLMs is that there is no architectural boundary between data and control paths.
But this combination of data and control into a single, flexible data stream is also the defining strength of a LLM, so it can’t be taken away without also taking away the benefits.
I tried this with early ChatGPT. Asked it to answer telegram style with as few tokens as possible. It is also interesting to ask it for jokes in this mode.
> But a lot of UK housing relies on on-street parking, and there's flats with car parks where charging isn't currently practical.
You forget the larger problem less wealthy individuals face: They typically already own a ICE-car and can‘t afford to purchase a new car multiple times in their lives.
But they pay off by keeping society clean and healthy and prevent loss of workforce from mental or physical disease?