An aside: Tim Harford presents one of the very best programmes on BBC Radio 4: More or Less [1].
Each programme investigates the reality behind statistics used in the media and by politicians. It's quite UK-centric, of course, but simply one of the most informative shows there is.
"By extension, this means that the most important factor determining success is the user experience: the best distributors/aggregators/market-makers win by providing the best experience, which earns them the most consumers/users, which attracts the most suppliers, which enhances the user experience in a virtuous cycle."
Sadly, the reality of the virtuous circle though is: enshittification.
"...from music to video to books to art; the extent to which being “special” meant being scarce is the extent to which the existence of “special” meant a constriction of opportunity"
I don't get this. Rarity or inaccessibility has been used as marketing tool, sure. But great music and books were not scarce for a long time pre-internet. Feels like a "never mind the quality, feel the width" view of culture.
"LLMs are breaking down all written text ever into massive models that don’t even bother with pages: they simply give you the answer."
When will people stop saying that? The give an answer, yes, but is it the answer: caveat emptor.
> But great music and books were not scarce for a long time pre-internet.
For some maybe. When I was first learning to program I would drive to the local bookstore and copy down code from books out of their very small tech section. I couldn't afford to buy any of the books at the time (out of the very small selection), and they were too new for the local library to carry. Now, I can learn about almost _anything_ for free within a few clicks.
Music was similarly gated, but more so by lack of money than overall access.
Previews only use the HMTL head - to get OG metadata - and usually an image.
Never thought about this before. One of my sites is (a single-page application) 45k HTML and 40k image. Hence about 50% of what is served for previews is wasted.
It would be nice if there was a way to recognise an "only html head needed" type of request. Don't think there is?
Nope, unless things have changed since I last dealt with it, Open Graph metadata is not considered "http-equiv" so does _not_ get served as an http header, only as a meta tag in the html header... so the HEAD verb won't get them for you.
The point that the leading group in a field could and indeed did know each other to some extent is somewhat conflated in the piece with that particular group, who represent a very particular set of one-off extreme high achievers.
Self evidently "most physicists" are not going to measure up against a group including 17 Nobel Prize winners.
PhDs serve a different purpose now in society, but are still of value.
I don't disagree with the article - enjoyed it. The 5th Solvay is somewhat of an outlier: majority of the folks in the famous photo were Nobel laureates or laureates to be, which must be some sort of record.
Unfortunately, everything is held back by the en(bull)shitification cycle we're in, not just academe. This seems to be an area where LLMs will make a big difference: what we think now is a flood will turn out to be the receding tide before the tsunami.
Is a "right imbalance", which seems to be the central idea expressed here, somehow being illustrated by a counter example: the content of the article itself?
- Natural Al estimated at 141.1 metric tons / year not varying.
- Human-made: 2016: 5.36 tons, 2022: 41.7 tons, "future" could be 912 tons, i.e. more than 6x natural.
I think the last figure is based on many planned large satellite constellations being deployed.