make sure to make your doctor a browser plugin to append think hard to all their prompts so you don't accidentally get the stupid chatgpt model diagnosing you

Yes

A photo of a scam advertisement on Twitter showing a fake CBC news article saying “The Whole of Canada is shocked after explosive interview: Pierre Poilievre says goodby to a normal life!”

Moving to Alberta isn’t thaat bad.

Tesla - Full Self Driving capable since 1901

That’s how we know there’s only one real Brodie Ferguson and the rest are imposters @smackcheck.bsky.social

So glad the very respectable and trustworthy people at the NYT have the power of bsky verification in their hands, so we can be reminded that the 'thought provoking columnists' they hire are credible and important speakers in the discourse.

A screenshot of a tweet from the user @The_Real_Fly. The tweet reads, "Trump didn't fold. He's pivoting after realizing his advisors were fucking retards. Now fire Lutnick and Navarro and we're back."

trump only hires the best people, except for everyone he hired in the first trump administration, who he hated and had to fire for being shit, and the new list we're working on right now for who is responsible for the incompetence this time, from the new people he hired.

We took passive investing too far. You can literally announce you're going to destroy the economy in 2 months, and people will say, "nothing I can do. impossible to time the market, it prices in future information" and watch drooling as the market does nothing, then crashes on economy crash day.

Type: Text Screenshot A screenshot of a tweet with the hashtag #TransDayOfVisibility. The tweet displays a photo of a paper pasted on a brick wall with graffiti. The paper's text reads: "THERE IS NO TRANS DEBATE THERE ARE TRANS PEOPLE AND THERE ARE FASCISTS WHO WISH TO DOMINATE AND ELIMINATE TRANS PEOPLE". The brick wall's graffiti contains overlapping shapes in pink, blue, and black. The tweet's author is "WuTanglsForTheChildren". There are 140 likes, 3.9K retweets, and 14.6K favorites. A pop-up box describing the photo is displayed as well.

Image of a tweet by Krassensteins reads "BREAKING: 23 States just sued Trump". There is also a pop-up in the center of the image that reads: "Type: News Media, Original Language: English, [速報] 23州がトランプを相手に、数十億ドルの医療費削減の決定を巡って提訴した。オリジナル言語は英語です。". The original language detected is Japanese which translates to "Breaking News: 23 states have filed a lawsuit against Trump over the decision to cut billions of dollars in medical costs. The original language is English."

Everyone loves talking about alt-text and AI on this site.

I 'vibe coded' BlueGoo, a browser extension so people can use free versions of Google Gemma/Gemini, to help make alt-text for their posts, or create alt-text for posts that don't have any.

Free, open source, multilingual, probably evil.

You should try outsourcing your replies to an LLM it’ll allow you to communicate more constructively and appear to hold nuanced positions with engaged critical thinking. Or are you purposefully replying poorly because you’re worried about copy-infringing an intelligent style?

Why would I need a note to virtuously use a tool?

Where is the line where running a model becomes acceptable?
I assume only taking an input prompt is bad.
1. What If I fine-tune a diffusion model?
2. Tweak tokenization + schedulers
3. Use control-net to get specific composition
4. Purposefully refine every element with noise masking

Okay, so photography can at least sometimes be considered art. Photography is using a tool and definitionally capturing already existing objects using it, plus post-processing with other tools. So clearly humans can create art using tools and pre-existing data. Which brings me to my question:

*jumping in a time machine to explain to myself 10 years ago*

'so in 2025, the right in the US will be working on expanding executive power to be an effective monarchy... and the left are the biggest, loudest and most passionate defenders of copyright and intellectual property rights.'

Is photography a valid art style? Only if you take a photo of something that has never been photographed before? What tools are allowed or not allowed?

As a thought experiment can you imagine the distant future where every style has already been created?

Am I not allowed to sing using my own vocal cords as every base melody has been created? Are we all to be slaves to those that came before us? Or do you support a time-based Disney style system?

What level of job augmentation are you against? Is using a shovel okay? What about a backhoe? A computer? Is excel acceptable? What about a compiler?

Using a computer to output copyright infringement or manually doing it should be the same thing.

So are you okay with using a computer to output AI art using a diffusion model so long as the prompt isn’t asking for something intentionally infringing like Ghibli style?

Do you think art styles should be off limits forever? Or, should they follow similar to a Disney prescribed copyright length? Should an artist born in the year 7000 not be able to draw anything because every possible style is owned by Disney-Ghibli corp?

How is using an opensource model weight on your local computer supporting them? They could go bankrupt, and I'd still have these 5-50GB files. You didn't answer my question on what level of human input you considered transformative. Why not use them in mindful and helpful ways? There's no going back

Where is the line where running a model becomes acceptable?
I assume only taking an input prompt is bad.
1. What If I fine-tune a diffusion model?
2. Tweak tokenization + schedulers
3. Use control-net to get specific composition
4. Purposefully refine every element with noise masking

Are you more pro-copyright than a mega corporation like Disney? Artists don't grow up in a vault never viewing other's work. If someone produces copy infringing work with GenAI it's no different than if they produce copy infringing work manually.

For this specific reply are you claiming that GenAI is not in fact best in class for OCR?

If you haven't been able to get anything of value out of GenAI that's a problem on your end. Looking at a set of incredibly powerful tools and not having the creativity or curiosity to get one ounce of benefit from them

The constitution is really inefficient. Doge could really optimize it, cut out the slop

My protest was about taxes on actual axes! I'm a medieval weapon enthusiast who never had a negative opinion on carbon pricing

Twitter intellectualism going well

2025: “Why is he refusing to discuss taking over Canada at the meeting”

My stablegenius bot is hauntingly realistic

Alright, ya friggin’ goof. We’re gonna have to treat ya like you’re from a rival province now, bud.