Hacker News new | ask | show | jobs
by hyperpallium2 1662 days ago
UBI will do this. Lots of freeloaders, but at pretty much the same cost as at present.

But it's instructive to look at communistic/socialistic states, which pretty much had this. Anecdotally, Joscha Bach talks about his father, being able to do his own thing in that environment, without needing it to be practical.

And perhaps that's the crucial thing: without incentives, ideas are not made practical, where they can make a difference. Did you ever notice that when some cool new mathematics is developed and applied to do something amazing, it turns out that the math had already been worked out by somebody else about two centuries ago - but that work had no effect on the breakthrough. It would have made no difference if it had never been done...

For me, who loves the idea of people being able to work on whatever inspires them, this is terrible news. I wonder if there's a way around it? Perhaps just better connecting previous work - "idea search", if you like (present academic "literature review" is evidently inadequate).

Perhaps a categorization system like Roget's or Dewey's, but for arbitrarily dimensional application of ideas, maybe somethig relational or like Hoogle for searching Haskell type signatures, which works surprisingly well, probably because types are general in terms of application. The semantic web doesn't seem to work well; too specific/concrete.

1 comments

> UBI will do this.

I get where you're coming from but UBI is not the same as funding individuals for the simple reason that funded individuals arrive at an amount greater than UBI funding -- economic firepower that's duly needed when competing for resources whose supply is far outstripped by demand (e.g. GPUs for academic compute competes with crypto miners).

True, though I think the most ground-breaking stuff is done by people without expensive equipment. e.g. Crypto algorithms themselces weren't prpven using expensive equipment. Tycho had the expensive telescope, but Kepler used his observations to make the breakthrough.

The reason is that if you must use expensive equipmemt within the ongoing process of your trial-and-error, it is not just expensive, but exorbitantly expensive.

Though... I guess youe example of- GPUs aren't that expensive, and can facilitate the reaearch process. OTOH I'm not sure there are breakthroughs being made that way, more iterative progress - the enabling breakthrough, of Deep Learning by Hinton, didn't need GPUs.