cheaty retweeted
asking claude "is this vibecoded dogshit with no real premise/reason for existing" and then uploading a complete mapping of my own brain
my favourite part with all new vibecoded software and ideas going viral is pasting the repo into claude and going
"is this vibecoded dogshit with no real premise/reason for existing"
and the answer is 90% yes, it is vibecoded dogshit
the other 10% is still usually impressive
0
4
39
1,000
do you think i can get ios running on this bad boy
(HTC G1, worlds first Android smartphone)
(HTC G1, worlds first Android smartphone)
4
0
22
1,399
this is my attempt at daxposting
my favourite part with all new vibecoded software and ideas going viral is pasting the repo into claude and going
"is this vibecoded dogshit with no real premise/reason for existing"
and the answer is 90% yes, it is vibecoded dogshit
the other 10% is still usually impressive
2
0
5
505
cheaty retweeted
@fffgrepw the world works now, everything is 'slop' top to bottom
you can push llms as a tool to make whatever the fuck you want without ever checking if a better alternative already exists, and these models are not proactive usually to push back on user's requests to build stuff or make recommendations before starting
this is why i feel a research/plan mode is still a good idea, but not for most work
i was able to build a fully native dx12 renderer on windows for an xbox 360 game 3-4x performant as the next best thing only because i built on other people's work, curating dogshit vibecoded slop from better vibecoded slop with real novel ideas for how drawcalls and render threads should be handled and other technical jargon, most people do shit from scratch because they have ai psychosis that the models are good enough to do this from scratch (they are not)
but you can push these models with the right skills and harnesses and research to point where the same model with the right info with perform 500x better than the one without
you can push llms as a tool to make whatever the fuck you want without ever checking if a better alternative already exists, and these models are not proactive usually to push back on user's requests to build stuff or make recommendations before starting
this is why i feel a research/plan mode is still a good idea, but not for most work
i was able to build a fully native dx12 renderer on windows for an xbox 360 game 3-4x performant as the next best thing only because i built on other people's work, curating dogshit vibecoded slop from better vibecoded slop with real novel ideas for how drawcalls and render threads should be handled and other technical jargon, most people do shit from scratch because they have ai psychosis that the models are good enough to do this from scratch (they are not)
but you can push these models with the right skills and harnesses and research to point where the same model with the right info with perform 500x better than the one without
0
1
8
430
my favourite part with all new vibecoded software and ideas going viral is pasting the repo into claude and going
"is this vibecoded dogshit with no real premise/reason for existing"
and the answer is 90% yes, it is vibecoded dogshit
the other 10% is still usually impressive
"is this vibecoded dogshit with no real premise/reason for existing"
and the answer is 90% yes, it is vibecoded dogshit
the other 10% is still usually impressive
5
1
43
3,066
my life after getting access to abliterated deepseek
what the fuck why wasn't i just doing this the whole time
glm 5.3 + deepseek v4.1 flash is an insane cybersec combo
thanks @bedros_p for putting me on this
10
17
829
17,670
APPLY HERE, i'm using @morphllm for defense and i'm very happy with them, even non-abliterated deepseek is veryyy fast
x.com/skeptrune/stat…
x.com/skeptrune/stat…
🚨 🚨 DEEPSEEK ABLITERATED REFUSES NOTHING & IS JUST AS INTELLIGENT AS THE REGULAR MODEL!!
that should scare the hell out of you. there's 0 tradeoff to deleting the weights that cause the model to refuse instructions. they have nothing to do with what makes it intelligent.
if you that contains any amount of sensitive user data then you should be running an abliterated model against yourself 24/7 to check for vulnerabilities.
it's now live on @morphllm with serverless token pricing. i think we're the first provider in the world to do anything like this.
you need to kyc & explain your intentions to use it. link here:
morphllm.com/privileged-acc…
0
0
12
1,145
lmao i made more progress with $2.5 and abliterated deepseek (thanks @skeptrune and @morphllm for the access) in 2 hours than i did with 2 days of frontier models because of their shitty cyber guardrails
goodnight
goodnight
11
5
204
6,021