HNHacker News
TopNewBestAskShowJobs

dontupvoteme

1,116 karma · joined April 2, 2023

when karma hits 500 i'll have to make a new account
submissionscomments
dontupvoteme··on Demystifying the Shadow DOM
I think css must be the worst invention in computing, at least all the obfuscated javascript i know is trying to do something (spy on me), instead css just bloats up 800ms on linkedin into the size of war and peace because of how much they throw their weird 6 character psuedohashes in (WHY)
dontupvoteme··on Demystifying the Shadow DOM
I still don't really understand the regular DOM and what has happened to web development since the good old days of geocities..

Why does everyone need a full operating-system tier browser to render any website with javascript?

dontupvoteme··on GPT-4 Turbo with Vision is a step backwards for coding
Despite all that I find GPT moralizes far more than Claude does. I don't think I've had a single complaint from it thus far actually..

Also it's a lot better at coding. GPT has become exceptionally lazy recently, but i consistently can get 500+ lines of code out of claude (it even has to spawn multiple output windows)

Perhaps the top end 4 might wrong slightly more clever code, but you're hard pressed to get it to do more than a dozen or two lines.

dontupvoteme··on GPT-4 Turbo with Vision is a step backwards for coding
Good thing Claude's a massive step forward.
dontupvoteme··on GPT-4 Turbo with Vision Generally Available
Yeah it's absolutely horrible at layouts.

I'm not 100% sure it's related but if you ask it to draw bounding boxes around things it's always off by quite a bit.

dontupvoteme··on Home insurers are dropping customers based on aerial images
Don't give aussies any ideas, they tried to make mathematics illegal. Sometimes they take the worst aspects of the british and americans and make it their own
dontupvoteme··on Why AWS Supports Valkey
They do that because it's required of them out of sheer necessity, but it's the absolute bare minimum they can do because capitalism dictates it so

Just like you might imagine a leech could evolve some attribute to help defend its host against other predators.

I can't speak on the elastic vs open search topic though - does being part of the fiefdom of amazon work as a survival strategy?

dontupvoteme··on Why AWS Supports Valkey
Why don't you compare tax returns to find out?
dontupvoteme··on Why AWS Supports Valkey
so trillion dollar public companies get to use all their legal rights (and lobby governments to extend them), but the little guy can get fucked if he does?
dontupvoteme··on Why AWS Supports Valkey
sounds like a problem for the financial industry, thank God they have money to pay for it! they certainly don't use any of it to actually support anything of course.
dontupvoteme··on Why AWS Supports Valkey
You don't know what restrictive means.
dontupvoteme··on Autonomous Overhead Powerline Recharging for Uninterrupted Drone Operations [video]
Can't wait for the failure mode where they sit over HVDC lines, fail because there's no alternating magnetic field, and then fall on them and hobble the infrastructure.
dontupvoteme··on World_sim: LLM prompted to act as a sentient CLI universe simulator
I know indirectly that claude is basically completely unbeatable in creative writing, even GPT4 can't hold a candle..
dontupvoteme··on World_sim: LLM prompted to act as a sentient CLI universe simulator
I ran 'destroy anime' and among the great lines such as 'Osamu Tezuka's astral form screams silently as he fades into the darkness', and 'Eliminating sacred geometry of large expressive eyes, spiky colourful hair, small mouths' it ended with "Homo Otakus is now extinct"

'delete techbros' also generates some great banter. you get some very dated and incorrect references though "Mt Gox fade into fever dream of capital burnt for naught"

Also surprisingly there, grammatical errors[!]

10/10, i can see why everyone has been using claude for their, uh, creative writing use cases.

dontupvoteme··on World_sim: LLM prompted to act as a sentient CLI universe simulator
That's very cool, I didn't know they did that.

Odd they call us `humans`, and also a bit that it doesn't say who it's named after.

dontupvoteme··on When Will the GenAI Bubble Burst?
IMO LLMs are basically the best thing since wikipedia for learning/autodidacts.

They're also pretty fantastic for discoverability if you do a bit of filtering to remove hallucinations. Only downside is you end up finding that project is abandoned for years or so.

dontupvoteme··on OpenAI: Start using ChatGPT instantly
Yeah GPT is incredibly lazy, ironically 3 is far better at not being lazy than 4.

I guess you benchmarked via API? I've heard even the datestamped models have been nerfed from time to time..

dontupvoteme··on OpenAI: Start using ChatGPT instantly
ironically the API is available most everywhere, and it has an ok-ish web interface (they're really heavy on prompt engineering it seems though)
dontupvoteme··on Can Demis Hassabis save Google?
I mean you can make any set of LLMs produce any set of output you want. The problem is that it isn't terribly efficient and you have to filter between the real stuff and the hallucinations.
dontupvoteme··on Can Demis Hassabis save Google?
What is it particularly good at? video at 1M tokens sounded neat, I've been very impressed with Cladue's 200k + pretty good reasoning + not lazy output at the moment.
dontupvoteme··on AI garbage is already polluting the internet
the pragmatic solution is simply to build your own, personal, dead internet.
dontupvoteme··on Nvidia bans using translation layers for CUDA software to run on other chips
So what happens when someone in e.g. China or Russia releases something who they can't practically prosecute, can everyone else then use it?
dontupvoteme··on Big Post About Big Context
The main thing I'd like large context for is the ability to shove a ton of API/Python/etc documentation into the window for a given problem so it knows better what to do.

In terms of the actual output it can be annoying if the convo goes on too long and it hits the window, but for technical things this hardly matters as you can just restart it - more of an RP issue as i understand.

dontupvoteme··on Big Post About Big Context
I'm not surprised, those documents probably contain a lot of information which causes the model to get confused/lost/etc.

It also seems to line up with random paragraph location doing the worst in that paper.

dontupvoteme··on The "End of Programming" will look a lot like programming (2023)
That is a bit silly as SQL is a language itself, it's not a thing that can write in (many) other languages while still understanding human natural language.

LLMs are that 'blurry jpg of the internet' but now also more finetunable/specialized.

I'm going to look at siccing some LLMs on old Jenkins errors today, ideally it will give me more time to spend on building things (and hopefully less fires later).

I obviously don't expect it to magically fix everything but I have a good idea of what they're capable of, what they need help with, and what their pitfalls are

I can now have some scripts and some "agents" scan through 250GB of logfiles and do various sanity test things and either report back or make a branch and try to fix it, before i even need to look at it. It costs almost nothing and if it finishes early I can have it "go do the same thing but in rolling 3 month windows to see wtf happened when"

FWIW I think LLMs are more powerful for the language side of things. I get a lot of the programming improvement from python/metaprogramming/etc - but the combination is like having a group of interns you can make do a bunch of tedious/shit work that nobody wants to do.

dontupvoteme··on Ask HN: Did you encounter any leap year bugs today?
We did but I only learned of it today! Completely broke a lot of systems until they set the day to March 1st and back
dontupvoteme··on The "End of Programming" will look a lot like programming (2023)
IMO it's being (an army of_ automated software engineer hitmen to persecute any regressions, bugs or even potential bugs your project might run into so that

1. things manifest a lot less

2. you don't have to worry about maintenance

I could see a proper structured LLM setup some day even resuscitating 'dead' software.

dontupvoteme··on The "End of Programming" will look a lot like programming (2023)
The fact that an LLM can code up the game of life/snake/etc faster than you could find a download link is an interesting thing.

If you give it some documentation, background info, code and traceback info (and know to avoid context window poisoning) you can find it's capable of even helping solve rather tedious CUDA/etc issues.

dontupvoteme··on The "End of Programming" will look a lot like programming (2023)
Have you met most programmers or used much software?

Consistency - This is what GPUs were made for, putting a bunch of 'effort' almost instantly into investigating any regression that any real SW dev would ignore because, life's way too damn busy.

dontupvoteme··on The "End of Programming" will look a lot like programming (2023)
Why not simply use AI to implement monitoring and evaluation so incredibly fine grained any human would go insane if they had to do it?
← PreviousPage 2 of 20Next →