domain:science.org
Dangerous professional voice is a superpower.
Can you elaborate? Is this simply bombarding bureaucratic drones with requests for specific documentation so you can create a paper trail for yourself on their inability to process something up the chain?
I actually was required to take Gallup's Clifton Strengths test for my business communications college class (at a discounted cost) which I actually found pretty good and a step above the other tests I've taken.
For example, one of my top traits was "Context". Basically, that I enjoy thinking about the past, like to think about cause and effect, etc. This is helpful for I think some obvious reasons, but also a weakness, because change can be tough, and it can sometimes slow me from looking for current opportunities. I think that's actually pretty spot-on for me, and at the same time it's not true of all people (a lot of people find the past boring) so it suffers a little less than some of the other tests from the generic-advice trap common to astrology and horoscopes. Most of the traits highlighted has some kind of pros and cons list, with the idea being to better understand yourself and to double down on what you're good at (and be aware of the blind spots for what you aren't good at).
Or, "Harmony" was another top trait. It can be helpful for sensing conflict beforehand, finding common ground, staying practical, etc. but also means that I might sometimes avoid conflict, seek too-easy band-aid fixes, or get stressed when people don't agree.
Now, I will say that it's oriented towards corporate-like utility, rather than some kind of 'accuracy', and any system of personality with cleanly separated domains with suspiciously similar numbers of sub-categories is a little suspect, but I also kind of like that aspect of it. Also it's identifying the top "strengths", but really it's just saying these traits are your strongest traits, somewhat divorced from if they are good/bad or on some kind of sliding scale. In that sense it's a bit more honest because it's not so much about "you are X category" but more "this blend of traits represents you best".
I don’t fantasize about being a warrior or a poet. I like democracy, and I think I live in the most peaceful, best time in all of human history. I think both of these fantasies are stupid.
idk, when I was super little and I would start acting up in public my dad would physically pick me up, carry me to the car, and say "we are never taking you anywhere ever again if you're gonna act like a brat". And I would usually shut up pretty fast after that. For in-house infractions they'd hide my toys or something until I calmed down. Seemed to work well enough. It's possible I was a more "mild" case though, because by the time I was in first grade I had already become a relatively docile teacher's pet.
Basically I'd throw out all the psychiatry shit and say "sink or swim kid, up to you". That's how people did it for, you know, all of human history up until the last ~50 years or so. You think they had L-Theanine 1,000 years ago? No they said pick up a fuckin' shovel kid or we're all gonna starve this winter.
I think we could all be diagnosed with a little PDA, yeah? I got PDA for days. I'm still a lazy piece of shit as an adult who doesn't like to do anything. The only thing that makes me actually acquiesce to the "demand" is a hard deadline (with consequences) and a swift kick in the ass. It never goes away, you just gotta learn to deal with it. People like me appreciate the kicks in the ass, trust me!
Yeah. You just gotta accept it and move on and make the best of what you got.
Honestly I think you probably could get it to work okay right now with current models. However, for something like this, you really need to have some above-average skills in prompting. You'd find it helpful to read something like Anthropic's prompting guide, although that one's specialized a bit more for Claude than OpenAI's stuff. Some of the advice is non-intuitive, and you might need some tweaking. For example, for Claude (has some unique preferences like wrapping sections in XML tags), they recommend something kind of like the following in terms of general structure, and yes, before you ask, order can matter. If you don't want to read through it, here's my abbreviated notes for a good prompt structure for something like this:
You are __. The Task is __ (simple one-sentence summary).
< context to consider first, including why the task is important or needs to be done this way. Yes, telling the AI "why" actually does improve model outputs in many cases >
< input (or input set) to take action on; at least for really long inputs, it should be near the beginning, short outputs this can go later >
< details on how to do it, guiding the thought process. This is where you'd put some version of your bullet points. Your layout seems reasonable but it's possible scaffolding or flowcharting a bit more explicitly, including perhaps what to consider, could help >
< explain how the output should be formatted, and the expected output (possibly repeat yourself here about the original goal) >
< optional: 3-5 diverse examples that help with interpretation of goals and reinforce style and formatting. Also optional is you could provide the thought process to reach those answers in each case, mirroring the logic already outlined >
< any final reminders or bookkeeping stuff >
Did you know that Anthropic actually have a whole tool for that process? If you follow the link, you can get a prompt generator (literally, use AI to help you tweak the prompt to find a better one), auto-generate test cases, etc. It's pretty neat. You can also somewhat mitigate confabulation here by adding a bullet point instruction to allow it to return "I don't know" or "too hard" for the more difficult cases. Also, it's possible that, depending on the level of tool use and thinking needed per bullet, that applying it to a giant music library would require some real money.
I will note that OpenAI's guide has some slightly different advice, but still pretty similar. The main difference is a lack of XML tags, and also, OpenAI recommends this structure:
< identity, style, high-level goals >
< detailed instructions >
< examples of possible inputs with desired outputs >
< context that might be helpful >
As you can tell, it's actually pretty similar overall. Yes, you have more control (as well as more complicated stuff to manage) when doing it programmatically via the API, but I think you could probably try via the normal chat interface with decent results. I should also note that if the AI doesn't need to use very much "judgement", you might actually do better with a well-prompted 'normal' model instead of a simulated-reasoning model.
They’re using fiber optics drones now which is insane
But how far away do you need to be to shoot out buckshot or something? Or netting? Drone kill zone is maybe 20 meters.
It does not appear that either side has figured out a way to hide from enemy drones. If you are in the settlement that they are attacking, you will eventually just die.
Given that they didn't remove any of those comments, that seems unlikely. You probably just got autojannied.
Vinland Saga: Suggest finishing, partly because the anime does something I've rarely ever seen: after the first season is a (fun) orgy of violence and revenge and action, the second season is the opposite: character development, slow plot, and a message about how violence is bad.
My own suggestions/mini review list, loosely sorted in order of appeal to non-anime types, or people who have only watched one or two. Of course, it's always subjective. I don't quite agree that "anime is just a medium" because its highly-controlled production pipeline and limited set of studios creates some definite commonalities, but it's true there's a wide variety of genres.
Violet Evergarden, 9/10
THIS is really an excellent first or early anime. A woman used essentially as a special-ops child soldier is now a little older, and while the war continues, she decides to take up an unusual vocation: a typist in an era where few people know how to write (I guess), she also assists in helping people organize their thoughts to write letters. Often, these letters are emotionally charged, or offer some major catharsis; thus the show's episodes are organized roughly with a major letter per episode. Parallel to this, we should mention that the main character, the eponymous Violet Evergarden, has lost both of her arms, replaced with mechanical ones, which mirrors her emotional state, still dull and robotic from her war experiences. So we slowly get to see her open up over the course of the series. Sad and emotional at times, hopeful in others, this one is highly memorable and at times honestly you often forget it's an anime at all. Finished, a season and a movie or two.
Apothecary Diaries, 9/10
This show is great. A nice mix of mystery, cool setting, and like the previous, much fewer anime tropes than your usual fare, this one stands out. A fairly level-headed girl but with a strange obsession with poisons, raised as an apothecary by her adopted father (read: herbal-medicine doctor for the poor, in this case often a brothel) is kidnapped into a loosely-Chinese imperial palace as a servant there. And not the cool, plot kidnapping version either, she's literally just nabbed off the street and sold and has to come to terms with her new life. Which she does, and she's pretty smart and a good investigator even though it really isn't her interest, and she gets pulled into harem politics to some extent as first a food taster, and then other adventures especially for a powerful eunuch within the palace. Ongoing story with two seasons, but with some good closure.
Frieren: At Journey's End, 10/10
Now, I'm not sure whether this score, which reflects my anime of the decade designation, translates to the general public, but it's very enjoyable. A fantasy series that explores the idea of what a long-lived elf's life is actually like! Lord of the Rings plays a bit with this idea in a way, but doesn't fully commit and it's spun differently. There, the elves are kind of tired of life, but here, we ask the question: what might Legolas be feeling, going on an adventure with some others, when he knows that they are going to die and leave him behind again? LotR dodges this a bit by both killing much of the cast, and Gimli is also of a similar long life, but Frieren tackles this a bit more explicitly. She once went on a save-the-world trip, but as the mage of the party. Living for at least a thousand years however, she doesn't fully appreciate the impact this trip had on her, and experiences regret for not emotionally engaging more after her friends pass away. She uses this as impetus to start another journey back north again to the demon lands, retracing the save-the-world steps with a new group of people who grow on her. The world-building is great, the storytelling is on point, the vibes are excellent, it's just a great watch. Ongoing, one season completed.
Kaguya-sama: Love Is War, 9/10
This one earns a rare distinction for carrying with it a strong piece of advice: try the first episode or two subbed and dubbed. A humor-first series, this one takes place in a super-elite high school, where the two highest-performing students (one old money, and one a scholarship student), on the student council together, are trying to get each other to admit a crush on the other. They play all sorts of mental gymnastics to make this work. The humor largely comes from the commentary/narrator, but the sub and dub both approach it differently (and the dub actually localizes many of the jokes, so they are funny but in a different way). The sub leans a bit more dry-humor, irony-focused, while the dub plays up the conflicts as being outrageous. At any rate, this one is just good fun and although the series starts out as a bit more like a series of connected skits, it eventually transitions a little more into a proper show with character arcs and plot and all that good stuff. Finished, three seasons, epilogue movie to come.
I don't really know how the below actually stack up but I felt like tossing them in too.
Angel Beats!, 8.5/10
Admittedly it has been a while since I watched this one, but it's good. Some nice emotional catharsis, but I don't know how much I can say without spoiling things too much. A guy wakes up all of the sudden in a sort of alternate-reality school, with a confusing 'war' between a group of kids within the school and the school student body president, who is a bit of a robot, with the drone-like other students as bystanders. Despite the presence of guns, this is a low-violence affair where the war is mostly a series of, well, pranks more or less? Despite the sort of confusing set-up, you get some good character moments, and this one is a tear-jerker at times.
Dandadan, 8.5/10
Visual flair. Panache. Dialing stuff up to 11. This anime is now in its second season and is ridiculous but fun. You probably only need to see a few minutes to get an idea about what this one's about, but for text purposes the classic hook is that a high schooler who believes in ghosts teams up with one who believes in aliens, and they're both right! He has his balls stolen by a spirit, and aliens try to kidnap her, and then they have some adventures trying to resolve that.
Code Geass, 8.5/10
This is kind of like the Ender's Game of anime in a way? The main character lives as a privileged elite in a dystopian Japan ruled by a world monarchy-autocracy, but decides to join a local Japanese rebellion. He's very much a 5D chess type of guy, takes on an alter-ego, and did I mention there's mechs for some reason? Most of the show is him outsmarting people, because in a similar kind of "hook" to the oft-recommended Death Note (which I personally don't like), he has the ability to brainwash-command anyone to do anything... but only once, ever, in their life. Which he obviously wants to keep a secret, but has to also be smart about using due to its one-time-use nature. Two seasons.
The Thick of It is like The Office (US version) in that it’s an idealized version of a ‘fun’ office as imagined by people from that particular culture. Leaving aside that even in the mid Blair era I doubt most of that kind of banter was tolerated all the time even from Alistair Campbell types (let alone random civil servants) there is an authenticity to it.
I would say that working in an office full of well-educated English people who like banter, at its best on Friday afternoons when everyone is comfortable with each other, has had a couple of drinks at lunch and is joking around then sure, it feels a bit like The Thick of It (at least to my foreign ears).
In the same way, Americans and some other Anglos identified with the kind of camaraderie and humor in the US Office because they experienced a lesser version of it, sometimes, themselves. The Thick of It lacks the maudlin sentimentality of most US sitcoms but a similar principle applies.
The swearing in particular seems like a remnant of the TV culture of that time, ‘The F Word’, Gordon Ramsey swearing, the growth of satellite TV without watershed, established networks being willing to have more swearing on later in the evening. This was, after all, when Little Britain was airing on BBC One. In addition, the main character is based on a notorious fan of profanity even today.
More interesting for the TV connoisseur is Veep, which while a less funny show highlights the subtle cultural differences between Britain and America by having American actors and characters speak dialogue clearly written by Brits and therefore always a little uncanny to American ears.
Succession (by much the same team) has a similar problem but skirts it by making the main cast half-English.
Tottenham is one of the worst parts of London, sadly.
My take is that competent people still exist, and there may be more of them than ever, but they've been pulled into niche industries where they can make much bigger salaries, leaving the dum dums to fill jobs in government and more mundane industries.
That was me, thanks for the rec! I'll definitely check it out.
People are unalived rather than killed.
I wonder if the bot would pass "mur-diddley-urdered". Deliberately make the censorship look stupid.
A girl challenging you and her showing you who's boss, and then doing that in the movie are different. Mark Renton in Trainspotting, when he meets Diane gets challenged by her despite his awful pickup, but ultimately she does not end up being the man in the movie. I like feisty women, modern portrayals, despite the nudity, are very sexless. You can sense that the creators have had weird lives; the woman running Star Wars, or whatever their latest flop TV show was, spent her high school as a pariah. No wonder her art is that bad.
Oh, a throwaway line about NPR hosts getting flogged.
I suspect my recent comment of the week about race and IQ to be the real culprit, but they got Capone for tax evasion.
Thank you! I'm currently at Borough Market, suffering from sensory overload. I'll see how my legs hold up after I've had something to eat.
pretty nice in summer especially if it’s sunny this afternoon
The opposite, unfortunately. Tottenham was positively dystopian due to the overall bleakness and the rather concerning number of schizophrenics on the loose. Clouds as far as the London Eye can see.
I have you pegged as "flighty wordcel who is way too interested in austere, self-referential literature and art" and that's meant as a compliment. The profile of your interests isn't super typical here and it adds flavour and depth to the Motte, I don't like it much when people downvote them.
Just seconding this. We need some nutters around the place to keep things interesting
Conventional English strongly associates quality with clarity, for good reason IMO (see C.S. Lewis and Orwell’s critiques of bureaucratese and the superiority of clear sentences and everyday Anglo-Saxon derived words, with which I largely agree).
But I do see the force of the Continental claim that writing something is not actually the same as expressing it. It can even be the reverse - an inoculation that robs an idea of all its true interest and allows you to lock it in a mental drawer without further thought.
The problem for me is that I find all the continental attempts to circumvent this process to be tedious in the extreme :) Which is why I appreciate having you around to try and indicate why it is not so.
where Peter Pan gets slapped by a girl in his own movie
My impression was that girls slapped protagonists a lot in their own movies in the Old Movies. It's what girls do.
Start in St John’s Wood or Primrose Hill, walk down through Regents Park, past the rose garden and the outdoor theater, down through Marylebone and some of Mayfair, then into Green and then St James’ park, then walk down along the Strand, stopping by whatever seems interesting. Then either take the district or circle line west, back to Hyde Park, to Chelsea and South Ken, past the museums (V&A if you haven’t been), all of Chelsea is pretty nice in summer especially if it’s sunny this afternoon.
Or walk down to the Embankment or Westminster pier, take the thames boat (now branded “uber boat” due to sponsorship but its regular public transport) to Greenwich, see the Cutty Sark and the naval college and the date line, have a pint at the Trafalgar, take the boat back.
Thinking and Feeling aren't so alien to each other.
Right, I'm always trying to explain this to people. The "logic vs emotion" dichotomy is clearly overly simplistic and not really tenable. But at the same time, I think it's pretty clear that different people do think and experience in fundamentally different ways, and we need some kind of language for talking about it, even if we end up not using those terms specifically.
I think the synthesis here is that we should have enough knowledge that if we were to build an ASI, and turned it on, it would in fact do what we tell it to, interpreted in the way that we mean it, and that this is table stakes for getting any sort of good outcome. - That is, our problem at the moment is not so much that we don't know what the good is as that we can't reliably make the AI do anything even if we want it very much and it is in fact good.
More options
Context Copy link