Tiina Meister

Karma: 1

Tiina Meister 18 Jul 2026 18:37 UTC
2 points
1
on: AI 2040: Plan A
I think there’s a problem underneath this that isn’t really about capability or timelines. Even today, entry level jobs are shrinking and students are collecting credentials and work experience from first year, before they’ve really learned anything, just to stay competitive. Nothing wrong with that individually, but it’s a race, and the thing it’s racing against is a machine that can eventually do the job better anyway.
The economy is already restructuring around this, not in some future scenario, now. Figuring out what a computer can do increasingly means figuring out what a human’s role actually is, not as an afterthought once alignment is solved, but as its own open question. Even in a good, safe, slow scenario, we still haven’t answered what a person’s role is once credentials and experience stop being the advantage they used to be.

Tiina Meister 18 Jul 2026 18:07 UTC
1 point
0
in reply to: Sherrinford’s comment on: AI pause: the case for ASAP
I do agree that it is hard to find right amount of regulation for AI, especially when it is such fast evolving and different countries use it – regulation goes down to values in the society – what are specific values that need to be protected. Take AI giving medical or mental health feedback: whether that should be allowed isn’t about getting the risk level right, it’s whether a machine should be doing that at all. Two people can agree on the facts and still disagree, because it’s a values question (what needs regulating).
Also, why does AI need to sound so human? For example ChatGPT says to me things like “in my experience” or “I’ve encountered several cases like this,” find it quite odd because it hasn’t experienced anything, it’s a computer. If we’re worried about AI overstepping into human territory (hence need regulation?), maybe part of the fix isn’t just regulating what it says, but asking why it’s designed to talk like it has a first person perspective at all. That shapes how much authority people give it and how it mentally shapes them, and that too is a values choice.
“Right amount of regulation” only makes sense once we’ve agreed what we’re protecting and whose values decide that. In a pluralistic society, I don’t think we can say we have only one truth? Especially when case is not black-white.

Tiina Meister 6 May 2026 21:49 UTC
1 point
0
on: Model Spec Midtraining: Improving How Alignment Training Generalizes
The findings are interesting. It suggests models generalise better when they understand reasons rather than just patterns. Which leads to question if explaining why a rule exists already helps, could the next step be abstracting across rules to find the underlying principles that generate them? Rather than teaching models a better map of rules, teaching them what produces good rules in the first place.

Tiina Meister 6 May 2026 21:19 UTC
1 point
0
on: Intelligence Dissolves Privacy
We spend enormous energy protecting bank statements at some level, treating them as surveillance risks. Yet people voluntarily share with AI systems something far more revealing: how they actually think, what they believe, what they are uncertain about, etc. A bank statement can tell you someone eats at the same place twice a week. But a long AI conversation tells you how their thinking drifts, what they are afraid of, what kinds of arguments move them, and thoughts they would never say out loud. I think the most sensitive data is no longer behavioural.

Tiina Meister 3 May 2026 15:36 UTC
1 point
0
on: Current AIs seem pretty misaligned to me
My take is that, current models are trained on enormous amounts of data and learn to produce what most people find good or useful. But when you ask them to create something genuinely new, they don’t really think, rather they recombine what they have seen most often, or at least this seems to me like that. The result rarely feels truly unique. I wonder if this is also why they tend toward mediocrity. If you train on everything, you optimise for the average. The average is not where the interesting work is.
Having studied law, computer science, and art history, I keep noticing that each field has underlying principles for evaluating whether work is good. How clearly is the idea expressed. How much noise is there. Whether the core idea is actually executed or just gestured at. When I moved between fields, these principles transferred easily, the fundamentals felt the same even when the surface looked completely different. Uniqueness in many fields usually comes not from inventing new elements but from combining existing ones in a way nobody has tried. Many architects work with light and shadow. For example S. Holl - I think because his idea is expressed more clearly and with less noise than most. The same seems often true in law, in music, in code.
If you trained a model to think in principles rather than patterns, it might generalise better across domains and produce less average output. It might also need far less data, if the principles are consistent, you don’t need every example, you need the clearest ones. Art history is interesting here because it already contains centuries of human judgment about which works lasted and why. That is a small and unusually well-filtered dataset. Although, i have no idea, whether it is currently technically feasible.