lostphilosopher

Someone could write a cyberpunk Three Body Problem with this premise.

In every discussion of AI eliminating or dramatically reducing the compensation for <some large double digit percentage> of “white collar” jobs (and probably “blue collar” too). It’s unclear to me what the end state is - the vast majority of the economy works on volume. You need large numbers of people with enough money to buy your product/service. As wealth concentrates there are fewer potential buyers and economies of scale start working against producers. (And governments need people with money to tax…)

I’ve been looking for tooling that would evaluate my prompt and give feedback on how to improve. I can get somewhere with custom system prompts (“before responding ensure…”) but it seems like someone is probably already working on this? Ideally it would run outside the actual thread to keep context clean. There are some options popping up on Google but curious if anyone has a first anecdote to share?

One challenge with hiring a nanny is if they need to take a sick day (or if they quit!) you can end up in a tough spot. In contrast a day care center usually has backups built in so you don’t end up scrambling.

Food pyramid was taught when I was in school, but that was before 2011 (as mentioned by another commenter) my own children are in school now and their school lunches align with more modern ideas (veggies and proteins). Certainly could still be improved but I recognize the cost, scale, delivery constraints, plus allergy considerations makes this non-trivial.

This seems like what should be a killer feature: Copilot having access to configuration and logs and being able to identify where a failure is coming from. This stuff is tedious manually since I basically run through a checklist of where the failure could occur and there’s no great way to automate that plus sometimes there’s subtle typo type issues. Copilot can generate the checklist reasonably well but can’t execute on it, even from Copilot within Azure. Why not??

I have had the experience of approaching or completing something potentially dangerous (merging on a busy street for example) and thinking I should “save” and sub consciously visualizing doing so internally. Very fleeting sensation and doesn’t happen consistently at all but it’s interesting when I notice it.

I don't think this is unique to churches, or even non-profits. Plenty of non-church non-profits rely on a few large donors for much or their funding (in fact plenty are designed that way out the gate - they're founded by one very wealthy individual to work on the projects they care about) and plenty of for profit businesses rely on a few large dollar clients for much of their revenue. Both could potentially be seen extensions of the same economic system that concentrates wealth for both individuals and businesses at top.

We don't really have a true test that means "if we pass this test we have AGI" but we have a variety of tests (like ARC) that we believe any true AGI would be able to pass. It's a "necessary but not sufficient" situation. Also ties directly to the challenge in defining what AGI really means. You see a lot of discussions of "moving the goal posts" around AGI, but as I see it we've never had goal posts, we've just got a bunch of lines we'd expect to cross before reaching them.

Related - there needs to be individuals and businesses that want/need and can afford upgrades and repairs. If office workers are getting replaced with AI we don't need to build and maintain offices and the ecosystems that support them (see also WFH/Covid) and those workers won't have income to pay for plumbers, electricians, roofers, etc. for their personal property. A worst case scenario AI workforce revolution would attack trades from both supply and demand.

It's worth noting that for those edge cases all the productivity monitoring in the world won't make that employee any more effective, and you won't need those tools to see that they're not cutting it (assuming you're engaged with your team as the other commenter describes). You'll likely lose more in annoying the rest of your team and burning your own cycles with surveillance than you'll gain from it.

I've long maintained that kids must learn end to end what it takes to put content on the web themselves (registering a domain, writing some html, exposing it on a server, etc.) so they understand that _truly anyone can do this_. Learning both that creating "authoritative" looking content is trivial and that they are _not_ beholden to a specific walled garden owner in order to share content on the web.

I don't know if AI books are or will be as good or better than human written, but to me this is the problem - "Even though artificial intelligence is still in its infancy, AI-made books are already flooding the market." There is no scarcity problem in books. There are already way more that I would enjoy reading than I will ever actually read. It's already tough to prioritize which ones to get to without having vastly more to sort through. And people _enjoy_ writing books. I don't want to support automating something away that people enjoy doing, is produced in abundance, and is very low cost to obtain already.

I was surprised at the time how cheap the original Pebbles were, they were nearly exactly what I wanted and I would have been willing to pay more for mine. In fact I ultimately paid more to replace mine with a watch I like less. When Pebble folded I wondered if having too low of price ultimately hurt them - if they didn't pick up enough customers to make up on volume what they left off the table on per-unit revenue? I hope the relaunch is successful, and I assume they have all manner of internal data that says I'm wrong, but my initial reaction to the listed prices is the same as it was to the originals - they seem too low. (I'm setting aside the caveat about a potential price change due to tariffs and assuming they launch at current list price.)

PhD itself is an abbreviation for "Doctor of Philosophy." The title is more about the original Greek "lover of wisdom" than about the modern academic discipline of philosophy. https://en.wikipedia.org/wiki/Doctor_of_Philosophy

Doctor is similar - in the US, when someone says "Doctor" they usually mean "Medical Doctor" but "Doctor" just comes from the Greek "teacher" / "scholar" which is more broad and the title can still be used officially and correctly for PhDs. https://en.wikipedia.org/wiki/Doctor_(title)

I’m less concerned about the supply of handymen inflating and more concerned about demand falling. If AI really does replace a huge portion of the people with the money to hire handymen and plumbers and the property to need them where do they get the work from? More people will DIY out of necessity even if the results are worse because they won’t have another option and fewer people will have the property to maintain in the first place.

Edit: hey, can your kids inherit your "digital content"? They can inherit your disc collection.

With arrival of the holiday season I brought out my Christmas CDs and records from storage. I use these exclusively for music in the house/car/etc., and part of why is that I have kids and I want them to be able to inherit these some day. I understand that physical media degrades and they may not be able to "use" these at some point, but they'll still have the objects and know exactly what versions they "grew up with" and could potentially track down / make replacements. (See also: https://www.goodreads.com/quotes/217710-this-milord-is-my-fa...)

I've had family members pass and I've appreciated having physical things I can hang onto, especially things like tools, music, and books where I can use/listen/read and feel a connection with them.

(Full disclosure I also prefer physical games, music, and books in general both for my own consumption and for ownership rights reasons.)

Even if X isn't the right solution to my use case I still often want to know _why_ X (or my implementation of X) doesn't work. The answer to that might be a really valuable learning independent of the problem at hand.

The Metal Gear Solid boss fight with Psycho Mantis is my favorite example of rarely-seen game mechanics (in this case in a video game).

(Spoiler alert on a game from 1998.) https://www.thegamer.com/metal-gear-solid-psycho-mantis-boss...

In the US I suspect "contract to hire" fills a similar function. I see that all the time but I don't think I've ever seen an official "probation period" on an employee role. (Could also be regional or industry dependent.)

"Show HN: I made a self-hosted UI for ChatGPT," perhaps?

Another possibility is that women are more likely to choose to self report their status as "in a relationship" vs. men. (Which would be an interesting finding on its own!) I'm not sure which source they pulled those figures in the first line from, but elsewhere it talks about PEW studies and the Survey Center on American Life which are presumably working from self reported survey data? "Married" is easier to track objectively than "single" vs. "in a relationship" which is more subjective. Is having been on three dates a "relationship?" Is it a relationship if you never talked about it with each other in those terms? If you're dating multiple people non-exclusively are you "in a relationship" or "single"? (I'm not asking; just speculating that two people might answer those questions differently.)

The 60% / 30% figure doesn't pass the sniff test to me.

Edit: Noticed the comment about this data focusing on "young" women and men and the possibility that more women are dating men outside that range than men are doing the same. That's also a good explanation. This was mentioned in the article too; I was just initially underestimating its significance.

I'm ok with not allowing multi "household" sharing in principle, but I doubt this can be enforced without causing problems for even for single household users and seems like an odd way to handle this. Thinking about travel especially. Restricting the number of currently streaming devices makes way more sense to me and should support "add a screen" up charges. Rather than fighting the "sharing" use case - why not find a way to monetize it?

I like the Laundry Files approach - James Bond-esque action plus Dilbert-esque bureaucratic and IT disfunction. (Also monsters.)

https://en.wikipedia.org/wiki/The_Laundry_Files

I think the increased prevalence of work from home is a huge factor for meet ups and urban activities in general. If I'm already out and about and likely nearby it's an easier sell to get me to go to a meet up. It's harder if I'm being asked to head to a central downtown just for the meet up, but there aren't enough interested people in my suburb to support a more convient meetup.

I'm excited for ChatGPT in video games. Especially in procedurally generated or Roguelike games.

Online booksellers often denote "mass market" vs "trade" paperback which does potentially give you info on the quality of the paper.

"The paper quality is better as well, as most often they are printed on acid-free paper like a hardback book." - https://biblio.co.uk/book_collecting_terminology/trade-paper...

(I've found this to be relevant when sourcing physical copies of out of print books.)

https://xkcd.com/722/

Magic for Humans with Justin Willman on Netflix (https://en.wikipedia.org/wiki/Magic_for_Humans) has some fun demos + explanations of this if anyone's interested.

What feels like a "random choice" when someone makes it is probably just the first thing they think of. So if you can make sure they're primed to think of what you want right away you can "predict" their choice. The real skill comes in being able to prime them without making it obvious that that's what you're doing.

Does anyone have a good explanation (or resource) for why this wouldn’t happen?