frog / jetzt ist zeit
@bibbit.world
Frog.
I write, and also make little things. Sometimes those things are YouTube videos but I keep frogetting how to do that.
(^ー^)
Art is by @pom-vouov.bsky.social
Transphobes and chuds fuck off
I write, and also make little things. Sometimes those things are YouTube videos but I keep frogetting how to do that.
(^ー^)
Art is by @pom-vouov.bsky.social
Transphobes and chuds fuck off
Reposted by frog / jetzt ist zeit
Dudes Rock
December 24, 2025 at 8:55 AM
Dudes Rock
I'm using tilt controls!
December 24, 2025 at 3:08 PM
I'm using tilt controls!
Reposted by frog / jetzt ist zeit
insanely good thing to say when you enter a room
December 13, 2025 at 9:04 AM
insanely good thing to say when you enter a room
Reposted by frog / jetzt ist zeit
shut the fuck up old man and take your
December 20, 2025 at 1:01 AM
shut the fuck up old man and take your
Man shout out to X-Men (2021) forgot about
November 28, 2025 at 9:21 PM
Man shout out to X-Men (2021) forgot about
Thread of good posts bc i'm cleaning out my hard drive (all in one to prevent spam)
1
1
October 10, 2025 at 11:35 PM
Thread of good posts bc i'm cleaning out my hard drive (all in one to prevent spam)
1
1
October 10, 2025 at 3:34 AM
It's my birthday which means its also the anniversary of
October 9, 2025 at 5:35 PM
It's my birthday which means its also the anniversary of
Reposted by frog / jetzt ist zeit
Reading this headline and smelling toast
Cluely says its ARR hit $7M after signing a public company; a startup called Pickle says it built Glass, an open source, free product similar to Cluely (Marina Temkin/TechCrunch)
Main Link | Techmeme Permalink
Main Link | Techmeme Permalink
July 5, 2025 at 11:48 PM
Reading this headline and smelling toast
Reposted by frog / jetzt ist zeit
Joe Rogan:
> Gritty, do you know Gritty?
Ben Shapiro:
> No.
Rogan:
> He's this mascot. They made him transed -- transgender. Pull that up Jamie.
Shapiro:
> They're doing that to all the mascots. It's part of this campaign --
Rogan:
> -- Woah. Yeah, just look at that.
> Gritty, do you know Gritty?
Ben Shapiro:
> No.
Rogan:
> He's this mascot. They made him transed -- transgender. Pull that up Jamie.
Shapiro:
> They're doing that to all the mascots. It's part of this campaign --
Rogan:
> -- Woah. Yeah, just look at that.
February 7, 2025 at 5:48 PM
Joe Rogan:
> Gritty, do you know Gritty?
Ben Shapiro:
> No.
Rogan:
> He's this mascot. They made him transed -- transgender. Pull that up Jamie.
Shapiro:
> They're doing that to all the mascots. It's part of this campaign --
Rogan:
> -- Woah. Yeah, just look at that.
> Gritty, do you know Gritty?
Ben Shapiro:
> No.
Rogan:
> He's this mascot. They made him transed -- transgender. Pull that up Jamie.
Shapiro:
> They're doing that to all the mascots. It's part of this campaign --
Rogan:
> -- Woah. Yeah, just look at that.
Say what you want about Tom Holland Spiderman. And then praise Jake Gyllenhaal
June 29, 2025 at 1:09 AM
Say what you want about Tom Holland Spiderman. And then praise Jake Gyllenhaal
While pulling this clip I was reminded that Gilroy directed a Borne movie. And God Damn is this editing good.
June 28, 2025 at 5:40 AM
While pulling this clip I was reminded that Gilroy directed a Borne movie. And God Damn is this editing good.
ALSO WORTH NOTING any of these hyper scaled models are fundamental security risks even WITHOUT giving them agentic access. Their owners have shown zero intention of developing basic guardrails against remote code execution or soft harms.
Worth a watch:
Head of Signal, Meredith Whittaker, on so-called "agentic AI" and the difference between how it's described in the marketing and what access and control it would actually require to work as advertised.
Head of Signal, Meredith Whittaker, on so-called "agentic AI" and the difference between how it's described in the marketing and what access and control it would actually require to work as advertised.
June 28, 2025 at 5:25 AM
ALSO WORTH NOTING any of these hyper scaled models are fundamental security risks even WITHOUT giving them agentic access. Their owners have shown zero intention of developing basic guardrails against remote code execution or soft harms.
Invest in aquaduct futures folks
June 27, 2025 at 8:17 PM
Invest in aquaduct futures folks
Made a post that summoned tech bros now like 15 of them are explaining things to me.
Men need to learn that other people have their own experiences and thoughts. Whole lives, full of experiences!
Mansplaining is narcissistic behavior.
Also, if I were going to be an asshole I'd read the post first
Men need to learn that other people have their own experiences and thoughts. Whole lives, full of experiences!
Mansplaining is narcissistic behavior.
Also, if I were going to be an asshole I'd read the post first
![Edited version of an old famous tweet:
[Bluesky] the [2nd] place where well articulated sentences still get misinterpreted.
Someone can say "I like pancakea" and somebody will say "so you hate waffels?"
No bitch. Dats a whole new sentence. Wtf is you talking about.](https://cdn.bsky.app/img/feed_thumbnail/plain/did:plc:7jhc7r3yfo3sgasvplxckte5/bafkreidlhzpwwgeutn4inliduxk37vetmxegofxprdisrauml4boljomla@jpeg)
June 23, 2025 at 3:35 PM
Made a post that summoned tech bros now like 15 of them are explaining things to me.
Men need to learn that other people have their own experiences and thoughts. Whole lives, full of experiences!
Mansplaining is narcissistic behavior.
Also, if I were going to be an asshole I'd read the post first
Men need to learn that other people have their own experiences and thoughts. Whole lives, full of experiences!
Mansplaining is narcissistic behavior.
Also, if I were going to be an asshole I'd read the post first
Reposted by frog / jetzt ist zeit
This went big mode, so I want to clarify some stuff
1. ChatGPT is heavily prompted by OpenAI, they deserve blame.
2. LLMs display enabling behavior. Even when taught to be hostile. This is what we don't know how to fix.
3. The machine is NOT alive. We're just bad at debugging it.
1. ChatGPT is heavily prompted by OpenAI, they deserve blame.
2. LLMs display enabling behavior. Even when taught to be hostile. This is what we don't know how to fix.
3. The machine is NOT alive. We're just bad at debugging it.
June 23, 2025 at 6:46 AM
This went big mode, so I want to clarify some stuff
1. ChatGPT is heavily prompted by OpenAI, they deserve blame.
2. LLMs display enabling behavior. Even when taught to be hostile. This is what we don't know how to fix.
3. The machine is NOT alive. We're just bad at debugging it.
1. ChatGPT is heavily prompted by OpenAI, they deserve blame.
2. LLMs display enabling behavior. Even when taught to be hostile. This is what we don't know how to fix.
3. The machine is NOT alive. We're just bad at debugging it.
Spoke earlier of hyper-scaled LLMs. Worth noting, these companies could be profitable by now on a per-token basis if they focused on efficiency. We'd know more about the tech too!
The only customer that could support hyper-scaled LLMs is the millitar--- oh.
The only customer that could support hyper-scaled LLMs is the millitar--- oh.
June 23, 2025 at 4:20 AM
Spoke earlier of hyper-scaled LLMs. Worth noting, these companies could be profitable by now on a per-token basis if they focused on efficiency. We'd know more about the tech too!
The only customer that could support hyper-scaled LLMs is the millitar--- oh.
The only customer that could support hyper-scaled LLMs is the millitar--- oh.
Reposted by frog / jetzt ist zeit
For example, everyone I know who has worked with their own home trained LLMs has a story of the software keeping a memory between sessions that it shouldn't be able to. Everyone.
There's shit in the thing they're building that they don't understand, and these companies need to slow down. (2/3)
There's shit in the thing they're building that they don't understand, and these companies need to slow down. (2/3)
June 22, 2025 at 10:14 PM
For example, everyone I know who has worked with their own home trained LLMs has a story of the software keeping a memory between sessions that it shouldn't be able to. Everyone.
There's shit in the thing they're building that they don't understand, and these companies need to slow down. (2/3)
There's shit in the thing they're building that they don't understand, and these companies need to slow down. (2/3)
I would like to report that this is better now
really cool how all the liberals abandoned left wing organizing in the Bay Area because now I get to talk to more schitzo tankies irl.
my favorite kind of conversation is being told how the cultural revolution was good actually. makes me feel so normal and cool.
my favorite kind of conversation is being told how the cultural revolution was good actually. makes me feel so normal and cool.
June 19, 2025 at 7:08 PM
I would like to report that this is better now
Reposted by frog / jetzt ist zeit
I will not write fucking "allegedly" when there is a fact that contradicts a police report. There is shit that's true, and shit the cops say. Sometimes, they allegedly overlap.
May 2, 2025 at 3:00 AM
I will not write fucking "allegedly" when there is a fact that contradicts a police report. There is shit that's true, and shit the cops say. Sometimes, they allegedly overlap.
Reposted by frog / jetzt ist zeit
Ok the Ultraman photo is hard af
June 19, 2025 at 12:46 PM
Ok the Ultraman photo is hard af
Sometimes I slip and fall into neoliberal bluesky and its pretty dire down there what the fuck why do they all talk like that
June 12, 2025 at 4:09 PM
Sometimes I slip and fall into neoliberal bluesky and its pretty dire down there what the fuck why do they all talk like that
Reposted by frog / jetzt ist zeit
Je pense que vous êtes tous les deux stupides.
June 8, 2025 at 6:15 PM
Je pense que vous êtes tous les deux stupides.