It will become harder and harder for the average person to gain from newer models.
My 75 year old father loves using Sonnet. He is not asking anything though that he would be able to tell Opus is "better". The answers he gets from the current model are good enough. He is not exactly using it to probe the depths of statistical mechanics.
My father is never going to vibe code anything no matter how good the models get.
I don't think AGI would even give much different answers to what he asks.
You have to ask the model something that allows the latest model to display its improvements. I think we can see, that is just not something on the mind of the average user.
Correct. People claim these models "saturate" yet what saturates faster is our ability to grasp what these models are capable of.
I, for one, cannot evaluate the strength of an IMO gold vs IMO bronze models.
Soon coding capabilities might also saturate. It might all become a matter of more compute (~ # iterations), instead of more precision (~ % getting it right the first time), as the models become lightning speed, and they gain access to a playground.
It doesn't solve the problem that people want what other people want.
In a "post scarcity" world we will figure out how to make certain things scarce and more desirable. Then people will start gaming the system to try to acquire the more expensive/scarce items. Some will even make it their life mission to acquire the intentionally scarce items/experiences.
The bigger issue to me is that not all geography is anything close to equal.
I would much rather live on a beach front property than where I live right now. I don't because the cost trade off is too high.
To bring the real estate market into equilibrium with UBI you would have to turn rural Nebraska into a giant slab city like ghetto. Or every mid sized city would have a slab city ghetto an hour outside the city. It would be ultra cheap to live there but it would be a place everyone is trying to save up to move out of. It would create a completely new under class of people.
> I would much rather live on a beach front property than where I live right now. I don't because the cost trade off is too high.
Yes, and?
My reference example was two aircraft carriers and 1:1 models of some fictional spacecraft larger than some islands, as personal private residences.
> To bring the real estate market into equilibrium with UBI you would have to turn rural Nebraska into a giant slab city like ghetto. Or every mid sized city would have a slab city ghetto an hour outside the city. It would be ultra cheap to live there but it would be a place everyone is trying to save up to move out of. It would create a completely new under class of people.
Incorrect.
Currently, about 83e6 hectares of this planet is currently a "built up area".
4827e6 ha, about 179 times the currently "built up" area, is cropland and grazing land. Such land can produce much more food than it already does, the limiting factor is the cost of labour to build e.g. irrigation and greenhouses (indeed, this would also allow production in what are currently salt flats and deserts, and enable aquaculture for a broad range of staples); as I am suggesting unbounded robot labour is already a requirement for UBI, this unlocks a great deal of land that is not currently available.
The only scenario in which I believe UBI works is one where robotic labour gives us our wealth. This scenario is one in which literally everyone can get their own personal 136.4 meters side length approximately square patch. That's not per family, that's per person. Put whatever you want on it — an orchard, a decorative garden, a hobbit hole, a castle, and five Olympic-sized swimming pools if you like, because you could fit all of them together at the same time on a patch that big.
The ratio (and consequently land per person), would be even bigger if I didn't disregard currently unusable land (such as mountains, deserts, glaciers, although of these three only glaciers would still be unusable in the scenario), and also if I didn't disregard land which is currently simply unused but still quite habitable e.g. forests (4000e6 ha) and scrub (1400e6 ha).
In the absence of future tech, we get what we saw in the UK with "council housing", but even this is still not as you say. While it gets us cheap mediocre tower blocks, it also gets us semi-detached houses with their own gardens, and even the most mediocre of the widely disliked Brutalist architecture era of the UK this policy didn't create a new underclass, it provided homes for the existing underclass. Finally, even at the low end they largely (but not universally) were an improvement on what came before them, and this era came to an end with a government policy to sell those exact same homes cheaply to their existing occupants.
> Some people’s idea of wealth is to live in high density with others.
Very true. But I'd say this is more of a politics problem than a physics one: any given person doesn't necessarily want to be around the people that want to be around them.
> If every place has the population density of Wyoming, real wealth will be the ability to live in real cities. That’s much like what we have now.
Cities* are where the jobs are, where the big money currently gets made, I'm not sure how much of what we have today with high density living is to show your wealth or to get your wealth — consider the density and average wealth of https://en.wikipedia.org/wiki/Atherton,_California, a place I'd never want to live in for a variety of reasons, which is (1) legally a city, (2) low density, (3) high income, (4) based on what I can see from the maps, a dorm town with no industrial or commercial capacity, the only things I can see which aren't homes (or infrastructure) are municipal and schools.
* in the "dense urban areas" sense, not the USA "incorporated settlements" sense, not the UK's "letters patent" sense
Real wealth is the ability to be special, to stand out from the crowd in a good way.
In a world of fully automated luxury for all, I do not know what this will look like.
Peacock tails of some kind to show off how much we can afford to waste? The rich already do so with watches that cost more than my first apartment, perhaps they'll start doing so with performative disfiguring infections to show off their ability to afford healthcare.
I use AI as a tool to make digital art but I don't make "AI Art".
Imperfection is not the problem with "AI Art". The problem is that it is really hard to not get the models to produce the same visual motifs and cliches. People can spot AI art so easy because of the motifs.
I think midjourney took this to another level with their human feedback. It became harder and harder to not produce the same visual motifs in the images to the point it is basically useless for me now.
I am a Bach fiend and the problem is BWV 1 to 1080.
Why would I listen to algorithmic Bach compositions when there are so many of Bach's own work I have never listened to?
Even if you did get bored of all JS music, Carl Philipp Emanuel Bach has over 1000 works himself.
There are also many genius baroque music composers outside the Bach family.
This is true of any composer really. Any classical composer that the average person has heard of has an immense catalog of works compared to modern recording artists.
I would say I have probably not even listened to half the works of all my favorite composers because it is such a huge amount of music. There is no need for some kind of classical music style LORA.
I don't question Bach's genius, but most baroque music doesn't interest me. Some of Bach's music I still enjoy enough that I could see myself listen to AI generated tracks made to generate music specifically similar to those pieces of Bach (and others) that I like. Though not enough that I'd seek out that in particular, and so I think the combination of what you say, with the general low-level of interest of those who would if it just happened to appear in my playlist still explains why it's not really a thing.
There are many artists, across the spectrum, like that for me, from geniuses that are just outside my own taste, to mediocre "one hit wonders" where I realise why they only had one hit when I listened to the rest of their catalogue but reallly would like more like that one hit (or handful)
And even when you like a broader selection of a composers music, there are time you might want "more of the same" of a specific piece. E.g. I quite like Beethoven, but I love the Moonlight Sonata, not just for what it is in itself, but the general systematic exploration of repetitive and slowly shifting of it.
There are other pieces by wildly different composers that invokes similar systematic exploration of patterns [1], but I'd also love to be able to hear more new "improvisations" over specific instances tuned very specifically to the aspects I like.
[1] On the extreme other "end" of these types of shiting repetitive patterns, I love Rob Hubbards Delta in-game theme of 11+ minutes of patterns repeated and iterated over as an illustration of the wide range that I like for much the same reason: https://www.youtube.com/watch?v=jOpIbm_XX-k
You have to assume the CIA are the absolute masters of layered deception.
I just listened to Anna Paulina Luna on Joe Rogan drone on about the CIA and remote viewing. I just assume that is all some kind of booby trap nonsense to fall into. I actually think the whole interview was Anna telling the bullshit the CIA showed her to keep her from finding anything that matters.
Same way with classifying the JFK assassination docs for decades even though there is absolute nothing in them.
It is brilliant. Something far beyond gas lighting.
Objectively, I have no idea what to believe with the CIA and that obviously is the strategy.
The JFK documents were classified to protect methods, and are still classified to protect people. People who are still alive, were tangentially involved (e.g. clerical staff or random bystanders who might've been interviewed but reported nothing of note).
It's why releasing more documents never reveals anything: it's stuff that wasn't worth trying to declassify because it's irrelevant, but it might contain a bunch of random names of people who are still alive and did things like sign for lunch that day.
My basic assumption is that were a top-level CIA officer or several involved in the JFK assassination, they probably wouldn't leave behind mountains of documents in a clear paper trail.
The problem with 2025 is I have seen thousands of better examples than that landscape. The reflections in the lake are complete trash.
Then I think of Veo 3 that is just incredible. So no, it is not impressive if a still from the video model is vastly better than the static image generator from the same company.
I find it especially annoying because I can't think of another company this would happen at. It is just so Google.
I think this is the gambit that we have already committed to.