4,357 karma · joined September 22, 2013
Software is also becoming more complex and expensive to build.
lolwutThe only way somebody could write this in 2024 is if they weren't around in 2004, or 1994.
In future posts, I’ll talk about what we are doing at Dry.ai to address this problem.
...or if they they're a marketing droid for another "AI" shovelware company. This is like the "there's GOT to be a BETTER WAY" segments of infomercials. come with a review about how they feel
I think you've got some misconceptions about what LLMs can do.You can ask one "write me a review for this movie in the style of Roger Ebert", but there's no critical evaluation happening there; getting "good" or "bad" reviews as output is a matter of tuning the prompt.
Maybe active funds would get more business if they were better at picking stocks instead of just whining about bogeymen like "regulation".
A round is run by using each function to calculate a play based on all previous plays, then comparing the results.
Now imagine you've got the "perfect" model as the computer player, that wins every time against humans.
A particularly crafty human player shows up, named Cantor. His function is defined as "the choice that beats what the computer's model returns". How often does he win?
This construction only works for simple games like RPS where the two players are interchangeable and a move has a well-defined "opposite".
Seems like complex functions could end up with effects lists of unwieldy size, similar to how Elm programs end up with a giant "Msg" type.
The only axiom is that smoothness exists, and in general increases with the number of observers
lolwutSome specific comments:
The "v" described is "the amount an observer actually knows about the system", apparently measured from zero to one.
"dt/dreality": what is this supposed to mean. It's not defined anywhere in the paper
er, "luxury submersibles", I mean :P
And now, ChatGPT does all that in seconds
I've seen ChatGPT generate a five-line Elixir function that had three hallucinated stdlib functions, several hallucinated options to the ones that existed, and several MORE hallucinated explanations for what those calls were supposed to do.IMO that's objectively worse than "copy-pasting from Stack Overflow" because sites like SO have at least some chance of obvious nonsense being moderated / corrected.
using an AI to go through all the files and do the math for me.
The only way you could rely on the the result would be to manually check each invoice, given all of these tools tendency to hallucinate. Might as well just manually collect the values at that point. I see no reason why this approach shouldn’t work perfectly in the coming years.
"I climbed the stairs two at a time today. I see no reason that I shouldn't be able to do the whole flight in one step in the coming years."Trying that with 95% hydrogen peroxide would work exactly ONCE, and "disinfect" the user entirely from reality.
https://git-scm.com/docs/git-log#Documentation/git-log.txt--...
novel sparse compute approaches
Unless "novel" is now a synonym for "not yet invented", this is jumping the gun with LLM-generated bullshit.This isn't "starting from an idea", this is starting from a Post-it that says "TODO: come up with an idea".
It will have all the code it needs to run and nothing more.
followed almost immediately by: There are plenty of multimedia frameworks that can be used to
create any video game or multimedia application.
Surely it would be EVEN MOAR EFFICIENT to reimplement just the exact bits of those frameworks that are needed, instead of loading the whole thing. It's probably more excusable to use an existing engine for 3D
This gets to the actual reason for the author's post: they think writing glue code makes them "better" than devs who use an engine. 3+ years including including obtaining the trademark itself
"Famous", eh? Must've been quite the ride over less than 4 years. I have strong claims for Federal Trademark Dilution and a
federal lawsuit for the same would quickly get me in touch
with counsel and I know they would resolve it in my favor
they’d be incentivized to drag their feet for the billables
commensurate with a federal trademark case
I see, you have a strong federal case but it goes to another school and we wouldn't know them /s millions of rows per device. After my where conditions
I know I can have around 100000 devices
That's going to be a fairly heavy bit of data; my wild-guess estimate is around 1TB for 100k devices with 1M rows each of 12 bytes (4-byte ID and 8-byte timestamp).Is there anything special about "specific_date_time_value"? An efficient architecture may be able to use that to avoid considering every row every time - for instance, using a queue to surface the "next" event easily.
So far I’ve tried modeling this in: - clickhouse - Postgres - mongodb
How did those turn out? What worked and what didn't?You've already collected a lot of info, sharing it will help!
Found a not-expired one on Reddit:
https://www.computer.org/csdl/api/v1/periodical/mags/so/2024...