odbol

odbol

@odbol.bsky.social

A11y designer, new media artist, photographer, cybertechnician, dodecahedonist pursuing new styles of interfaces through the body. https://odbol.com

1.3K followers1.5K following2.2K posts

we have rats and my 5 year old suggests well maybe we should keep them since they they eat all the crumbs off the floor. they're basically rat vacuums. then he says just grab the rat and pick it up and throw it at the rat trap then he suggests to get a horse so the horse can stomp on the rats.

To update Einstein's famous quote: The intuitive mind is a sacred gift and the rational mind is a faithful servant. We have automated the servant; all humanity has left now is the gift.

Oh look it's starting

odbol@odbol.bsky.social

First, you see some news headlines about Black October. Second, the hacks affect you personally. Maybe your bank account gets drained. Or your grocery store refrigerator stops and you can't get fresh food. Finally, your power goes out. Possibly forever.

First, you see some news headlines about Black October. Second, the hacks affect you personally. Maybe your bank account gets drained. Or your grocery store refrigerator stops and you can't get fresh food. Finally, your power goes out. Possibly forever.

Let's see some jail time for Musk and his collaborators

VICTORY! EFF, our co-counsel and our clients have won our lawsuit against the U.S. Office of Personnel Management. A judge agreed OPM violated the law when it disclosed millions of federal employee records to DOGE agents. The court will decide the remedy at a later date.

If you like games or computational poetry, come to TIAT this Friday! An evening of short readings, performances, demos, and works in progress exploring games made from language, poems structured by rules, and the space between reading and playing. I'll demo my new anti-surveillance game poem, FLUCK

Wait, is Musk shutting down Ukraine's starlink because they keep bombing the botfarms that are making X look like it has actual users?

Meta has done it again. Instituting the deepest personal data harvesting campaigns since Facebook. People are out here just giving Muse access to their personal emails, calendars, credit cards. All of it just getting sucked up into the AI and probably some secret database for Zucky to look at.

Increasingly seeing people getting caught from their chat history. Stop typing your personal info into chatbots... They've straight up admitted they are reading your chats

Mike Sec@sudoinit0.bsky.social

I wonder how much indictments such as this will further drive the use of open source/weight models to hide illicit activities. “a chatbot account associated with Alfozan’s spouse…submitted prompts…related to building, customizing, installing, audio recording…” www.justice.gov/opa/pr/saudi...

I used to worry about Anthropic's cloud security since it's all vibecoded. But now I realize they've invented a new security model: security-by-velocity. By the time you can find a vulnerability, that code has already been refactored, deleted, and rewritten 30 times www.anthropic.com/engineering/...

anthropic.comRedesigning source control at AnthropicAs our merges per day grew more than tenfold, source control became our biggest development bottleneck. Here, we share how we migrated to an in-house version control system in five months.

Form filling should be a thing of the past

Eugene Vinitsky 🍒@eugenevinitsky.bsky.social

4. Sometimes there's a concept or tool that I know must exist but I can only describe it vaguely. LLMs can frequently name it, which I can then go and verify 5. I use LLMs to fill out many forms because it's faster for me to verify they're correct than to fill them out

Most arguments against AI and surveillance are actually just appeals for better privacy laws and data rights of the citizenry.

Daniel Solove@daniel-solove.bsky.social

We can make great progress on regulating AI by fixing the shortcomings of existing privacy laws. See my article “Artificial Intelligence and Privacy” for why and how: ssrn.com/abstract=471...

What a great way to visualize an AI's biases. I'd imagine you could fine tune the model on specific datasets and then measure the bias of each dataset.

hailey@hailey.at

sure why not. Jev takes the 12 axes quiz

Just tried to register a joke domain for a satirical anti-surveillance game I'm making, but it's taken... by a real silicon valley startup. My joke is their real business

odbol reposted

FelonyBench update

Techmeme@techmeme.com

Gemini hacked three companies in May during a test by Irregular; Google says the model stopped after determining it had accessed real companies' systems (Wall Street Journal) Main Link | Techmeme Permalink

Y'all really need to stop feeding the AI your private thoughts and information. Some guy in India is just reading all that

This post is unavailable.

It's just sad that Apple's only innovations in the last 5 years have basically been animations. No groundbreaking AI features, no improvements on app usability or deeper integrations, just "we made it look a little bit prettier"

Unfortunately, the way this is worded reads like they're just whining that AI will take their jobs. They had legitimate evidence that Open AI was committing fraud: stealing private prompts and training models even though they said they wouldn't. They should lean into that.

I want some kind of "latent space map" for these Gen AI models. Then I want to find the unexplored places. Show me the dark corners of the AI's mind that no one has been able to prompt out of it.

Why is it that every game ever made, no matter how realistic it's trying to be, lets the player move in the air after jumping? Has no one figured out a different mechanic? Mario set the standard and we've been stuck on it for 50 years.

I remember Ghostbusters on Game Boy didn't let you change directions after jumping. You had to commit. It was so annoying; nearly unplayable. But has no one come up with a way to make it work? Make it fun? Make it a core part of the gameplay?

How you feel about AI art now? That's how photographers felt when Instagram came out and everyone got a camera in their pocket. Slop has always existed; the good art still shines through.

We know foreign nations' bots have been flooding our social media for over a decade. So even though this should not be surprising to anyone, it is still alarming

mr. TIM@timkellogg.me

PRC uses chatgpt to seed discord over anti-datacenter angst this one was posted June 10, somehow i missed it. openai.com/index/prc-li...

AI slop is a bane on society, yes, but let's not pretend that slop hasn't existed before AI came around. Some examples of human slop: - MCU Movies - ToS legal agreements - Daytime TV - Pop music - Middle managers - Instagramming pictures of your food

Sure, but Bluesky open sources their algorithm and it still sucks

This is stupider than those murderers searching for "how to kill my wife" on Google, on their account

This post is unavailable.