HN user
i'm sorry!
hn_serf@gmx.com
Contrary to the common myth, thinking hard does not burn significantly more calories. (Demonstration: Start thinking hard, and watch as your heart rate does not increase.)
this is wrong [0] , it's just hard to measure.
yeah, but the reality is that every chat room bot integration is half-assed, and the good integrations force you to use chat rooms/services that are either a nuissance to spin up and join, heavily opinionated, or lack features, or even the most basic security and privacy stances.
there's definitely room for someone to do it right.
i'm skeptical if it'll be dorsey, however.
the most obvious display of risk aversion is the greenlighting and production of "Spiderman 35 Electric Boogaloo" and "Batman # 740 The Dark Knight Re-Rises Again" -- meanwhile indies are still struggling to find first investors or writing money.
as I said in another comment : These large firms do not take risk.
all one has to do is look at movies from indie groups versus conglomerates in order to pick apart every point you made aside from technology investment and movie volume.
your most laughable point is the one about risk : no one takes less risk than a huge conglomerate beholden to a board and shareholders.
real risk in cinema comes the cast and crew that decide to forego dinner in order to get their vision to the screen, and the arts are filled with such passionate examples.
Not everyone wants the 1500th star wars or marvel movie, even the fan bases for those releases are tired.
there's a rare form of that where it goes from( simple -> programming language -> language with DSL support -> simple dsl ) ; emacs has followed this model a bit.
that's a real frustrating form of scrolljacking, a forced timed animation & lockout for concurrent scroll events until the timer is relinquished and input ceases; impressive.
for me it's often codebase decomposition.
it's hard to modularize a monolith without keeping large chunks in context at first pre-delegation; the orchestrator -- however you implement it -- needs to carry as much of the original thing as possible into context and big monoliths make this task heavier.
so, tl;dr : I use a lot of tokens re-writing legacy codebases that were constructed with very little CS training in some huge spaghetti fashion by random people around the world.
I use GLM5.2 and gpt-5.5 and 5.6 and never noticed any quality degradation near the limits of the context or due to compaction.
that's because you and your workloads probably fall into a bucket of users where compaction is tuned well.
but the reality is that those processes need to pick what to carry over, and that's a tough thing to get right for everyone, so for me compaction is a signal to restart the session or to tell the LLM to reference existing design docs lest we go off the rails.
large context limits is one of the things that make the open model competitors very attractive to me.
we don't have master AI that can create a great product out of a mediocre prompt.
but we do have sufficient AI to make a great product out of a great prompt.
garbage in -> garbage out hasn't gone anywhere.
so: much like to anyone that blindly complains that their compiler hates them : if you actually want help, provide information. If you just want to complain that the compiler is mean, scream at the sky.
plenty of people have figured out how to get this to work; more than enough to confirm that a straight <gambling-machine>/<hallucinatory-psychopath>/<random-number-generator> analogy is too simplistic to explain what we're working with.
it's not as hard as you're making it seem, you just allow for a positive or negative time offset and rely on the devices counting seconds similarly, which is mostly a given now.
I did this with raw footage and a VESC speed controller dump to overlay a bunch of motor stats on a prior project, this was pre-AI and it was still only an afternoon project.
ffmpeg/image-magick do all the real heavy lifting out of the box.
the site almost sells you on a ford maverick
"Wow, the same style engine, a reputable dealership network, a hybrid system with battery, and a turbo charger for only another 7 grand?"
This article reads like a description of gambling-addiction behavior; the author appears to be addicted to using LLMs.
you're projecting, it's just insanely irritating to work with a tool that 1) limits its' own use 2) with a random interval.
keeping in mind that plenty of people are making money on token use..
this guy sets a timer to wake up for work; he appears to be addicted to work.
why would we assume a corporate entity wouldn't manipulate popularity numbers on an owned and unregulated platform that sells their own goods?
https://80.lv/articles/former-valve-developer-claims-steam-l...
owning 'the list' for a thing makes a company, it's why billboard still has any relevance.
If it was just recognizing license plates that would be fine.
not really.
I don't want any private entity to catalogue humans identifiably to a population-scale extent.
if we as a collective whole wanted these things around they (Flock) wouldn't have to buy their positions from the regional municipalities.
(token usage) is trending towards predictability for a lot of reasons. it's not deterministic but it's getting easier to reason about usage.
this was released in 1996.
Microsoft was at one of its' most powerful evil phases it had ever seen during that phase, and to pretend it was some kind of antithesis to 'corporate metric please' is a disservice to history.
I liked comic chat , and I see that your actual point is more just "ai bad" , but 88-99 microsoft was brutally corporate metric pleasing.
see also : Microsoft antitrust history Microsoft FTC investigation 1990 Microsoft DOJ antitrust 1993 Microsoft 1994 consent decree Microsoft anticompetitive licensing Microsoft per-processor licensing Microsoft consent decree Judge Stanley Sporkin Microsoft vaporware antitrust Microsoft market foreclosure 1990s Gary Kildall Microsoft controversy Stac Electronics / DoubleSpace Microsoft Stac Electronics lawsuit Microsoft DoubleSpace patent infringement Microsoft Intuit acquisition antitrust
feels like selling an old bicycle on craigslist with the amount of things you can tag M$ with.
I don't know about the other industries, but within old cycling circles they're also just seen as the group that was first to make good counterfeit goods that were actually high enough quality to use, and their entrance to the market was aided by the fact that existing players, campagnolo and schwinn mostly, were producing junk from junk proprietary alloys in order to compete with 'overseas' components at truly exorbitant costs.
that was 40-50 years ago, so who cares, but Shimano is truly interesting for cultivating so many different opinions.
I think 'enforcement' is wholly outside the ballpark of a memory system.
man it sucks that waveshare doesnt sell a 7 incher w/ integrated esp32 like they do with every single other round display, could've had the whole package ready to go.
it's dead accurate for you because it aligns with your ideas on the subject.
if it had been for defense or some other such thing the media outlet wanted to celebrate it'd have been worded "Our defense apparatus sips only 23% of the country's power!"
tl;dr : accuracy isn't what aligns to personal interests, and journalists who choose to use language like this are (generally speaking) disinterested in accuracy and honesty as top priorities.
it's estimated around 5% of people remained covid-free & vaccine-free. I guess it's an easy guess that they would've caught it, but it's not certain.
a new hot take spotted : newtonian physics isn't a thing.
let's see how well the freeways work once we stop cleaning up after the accidents.
You don’t want to hire that guy. It’s not worth it.
how many times in history does this need to be disproven? some of the most insufferable assholes in human history were absolutely brilliant once-in-an-era types.
it's a chicken-or-egg question to ask whether or not the brilliance comes from the confidence or the other way around, but opinionated assholes that can't operate in traditional team settings have a proven track record of effectiveness in the world.
the trick isn't to ignore the brilliance, the trick is to find a handler or method that works to smooth the burden so that your organization can take advantage of the brilliance without self-destructing.
in fact, here's a rule of thumb. if any opinion anyone ever concludes with "Let's disconnect this human from society until they fix themselves" , well , I humbly disagree.
there isn't a market for barebones for the sake of privacy, but there is a market apparently for 'barebones for the sake of value', which is what the whole Slate truck thing is attacking.
I'm sure a chinese EV group could key in on the same pure-value market if there isn't a group already doing that. 'Golf carts for the street.'
such a cop-out when its policy shifts that are destroying it.
let's be real : it's microsoft.
yes?
it says it uses the haptic motors like within 2 or 3 sentences, it says it 'slams the controller into the puck until it charges', it says it uses an overhead camera, and it even says that it navigates the controller to the puck using the motors and camera further down if you don't feel like taking the 3 seconds to connect the dots.
a video would've been nice, but I would've communicated this the same way and i'm not an llm, even if this thing may have been written by one, nor did I have any problem understanding the intent - not a brag, it just doesn't seem that strange to me.
in pecunia veritas , an old HN favorite.
the things that empower modern toy drones were export restricted for years before hand.
I have seen this pattern over and over again. Anytime someone is an actual expert at anything, AI output appears insufficient or incomplete or outright misleading
media is awash at the moment with experts chiming in to support AI, saying their fields are being revolutionized, etc.
it seems unsurprising to me that the laymen opinion would follow the loudest media trumpets.