Neat! The nostalgia is palpable.
Off topic: we always referred to mode-x variants as MCGA, TIL that's more of a colloquial term than a technically accurate one [1]
[1]: https://en.m.wikipedia.org/wiki/Multi-Color_Graphics_Array
448 karma · joined April 3, 2011
Off topic: we always referred to mode-x variants as MCGA, TIL that's more of a colloquial term than a technically accurate one [1]
[1]: https://en.m.wikipedia.org/wiki/Multi-Color_Graphics_Array
kind of organically landed on a similar thing, where initial task is kept pretty simple (but not covered by common libs)
Some folk solve and code this quickly, so we add a requirement. (Generally up to a few stages - think like Advent Of Code)
Any leftover time can be spent discussion wishlist items for productionising, optimising, tests, etc..
But: plenty of people only complete the first stage, but this can absolutely be a pass too (based on comms style, approach, etc..)
Yeah spot on - exactly what I was aiming at too...
Having the intuition to use reverse (& pragmatism to keep it simple) is a great sign, and realistically more valuable [to me] for hiring than implementing rudimentary algorithms, but... still want at least a bit of both abilities
Also knowledge of standard library / language features is useful but (for me) doesn't equate to abilities in complex problem solving.
But honestly could be bias :)
Search landed on this neat summary [^1] (2024) from the lab(?) which also has a link to the original paper [^2]
[1]: https://www.quantinuum.com/blog/quantinuum-scientists-have-p...
That said: if I ever gave the palindrome question to a candidate (I haven't) and they pulled out a `reverse` library func/keyword we'd be encouraging them to do more.
Just because your project might not be at Google's scale doesn't mean it is therefore also not complex [^1]
Example: I'd say plenty of games fit the author's definition of "complex systems". Even the well-engineered ones (and even some which could fit on a floppy disc)
[1]: https://en.m.wikipedia.org/wiki/Affirming_the_consequent
> His parents, upon his graduation, had decided to give him a monthly allowance of $200 out of their earnings from Cobblestone Gardens, a substantial sum in those days.
> It was enough to keep him going, and indeed it guaranteed his survival for the next twenty-five years, arriving with welcome regularity. The allowance was a ticket to freedom; it allowed him to live where he wanted to and to forgo employment. [1^]
And $200 in 1937 gets you a fair bit of freedom (roughly $4500 in today's money) - sure, it's not private jet money but notable. And would have gone pretty far in Mexico and Tangier.
[1]: https://en.wikipedia.org/wiki/William_S._Burroughs
[2]: https://www.in2013dollars.com/us/inflation/1937?amount=200
and all those trips to Interzone don't pay for 'emselves!
All grumpiness has vanished (Though a modicum of despair remains)
Does seem to mostly be "here are some people in touch with current developments happening in the world thinking about their next phase" more than full-blown futurism - but still super interesting to read (if only to see the vestiges of tech optimism pre-dotcom-enshittification)
A few things which leap out that weren't included in the original post:
- "everything will be smart - that is, responsive to its external or internal environment"
Predicted the glassy eyed dream at the time, but not the more cynical reality we got where that smartness more often benefits the seller than the consumer
---
- "in-depth personal medical histories will be on record and in full control of the individual in some form of medical smart card or disk"
Another fun combo of rarely safe prediction based on developments of the time mixed with (in hindsight) overly optimistic assumptions about commercialism
---
And the bit summarised as purportedly predicting "(the internet)" actually has a pretty neat bit which was missed
"face-to-face, voice-to-voice, person-to-data, and data-to-data communication will be available to any place at any time from anywhere."
Predicting Skype/zoom seems a tad more interesting (even though things like Dick Tracy or 2001: A space odyssey had long had us wishing for it before 1995)
The preceding bit aged less well
"...broadband network of networks based on fiber optics; other techniques, such as communications satellites, cellular, and microwave will be ancillary."
Interesting that they appeared to be emphasising an _increase_ in wired data transit
But not surprising, as back then any cellular comms were far from impressive in comparison to the big-ass wired links you could use from universities
(That said, even back then we had microwave links acting as secondary high speed links for campuses which didn't yet have fibre or coax installed)
I was using the internet in 1995 - and pretty sure at least part of that connection use was over fibre, so I don't imagine someone trying to get away with predicting that.
But the path to my house was still a phone line - so I assume the original prediction was more about that fibre eventually running to our houses - certainly something many of us were looking forward to back then particularly any time someone in the house picked up a phone and killed our connection :)
(EDIT: have added an amendment in reply to this comment after reading some of the original PDF)
---
As for high contrast digital paper, that was also already part of the collective wishlist
Neal Stephenson's The Diamond Age [1] came out in 1995 and featured an e-book with an embedded AI and nanotech based display surface
BUT e-ink was first investigated at Xerox PARC in the 70's and was already under development by 1995 at MIT's media lab (commercialised in 1997 but hardly as bug a leap as it sounds) [^2]
[1]: https://en.wikipedia.org/wiki/The_Diamond_Age#Allusions_to_T...
Also - Honestly don't even understand why so many people would need to scrape for training data.
Is it naïveté (thinking it necessary), arrogance (thinking it better than others, and thus justified)?
Aren't most advances now primarily focused on either higher level (agents, layered retrieval) or lower level (eg. alternatives to transformers, etc.. which would be easier to prove useful on existing datasets)?
Genuine questions, all of these - if I'm off the mark I'm keen to learn!
Would be nice to see some effort on this front, a la "we don't scrape", or "we use this reputable dataset which scrapes while respecting robots.txt and reasonable caching", or heaven forbid "we train only on a dataset based on public domain content"
Even if it was an empty promise (or goodwill ultimately broken) it'd be _something_. If it exists, I'd certainly prioritise any products which proclaimed it
Do we need to make one?
I probably wouldn't even use it, but don't like to reward stingy dark patterns
https://clearthis.page/?u=http://fweet.org/
(Not affiliated with either site, and haven't done due diligence on clearthis - was just first site I found which appeared to behave as expected)
Still haven't personally dug through Joyce's notes (as RAW clearly did) but just getting to see the book through eyes that have was mind blowing in my 20's
Many thanks, kind stranger
Might be worth including a privacy statement about the handling of people's handwriting (meta)data, ie. disposal, not selling or using it as training data (assuming that's the case of course!)
No idea how many people would care, but was the first thing I looked for (habit before I share new things with non-techy friends)
Good to have linguists around reminding us to boldly go explore :)
Did you discover anything interesting in the process of building it?
TLDW: some fascinating realignment of perspective from a linguist about such things [^2], and the - frankly moderately mind-blowing - revelation that despite caring about language I've been gleefully pronouncing "covert" 'wrong' my whole life (and will likely continue to do so)
[1]: https://youtu.be/I6duEGj04Mg
[^2]: edit for clarity: such things being the concept of "normal forms" or rules in language.
Not nearly enough bespoke work going into building such forward-looking technology standards
Best of luck with it
Does seem like a trend too - not sure it can be attributed to headline brevity
Looking forward to having full-brain time, super intriguing to see folk so much further down the eBPF tracing path
Best of luck with the offering!
Alas (OFF TOPIC) clicking the burger menu on the site (mobile) gives a rather novel fullscreen black error
> "Application error: a client-side exception has occurred (see the browser console for more information)."