really love to see this!

Playdate wasn't on my radar until this, but it looks really cool and I'm into supporting projects that reject AI so I'm thinking about getting one

they've been pronouncing the "atelier" (in "Witch Hat Atelier") as "I tell ya"

(it means workshop and is pronounced more like "at-ell-yay" - source: intermediate french learner who does not pretend to be an authority!)

this is so vindicating, i love it so much lol

gpd win mini 2025 with screen showing RustRover. code on one side, mistral cli on the other with info about heap vs stack based data types.

mistral admin interface showing $0.67 usage for the month of april. (paying for per-request api usage instead of monthly flat rate)

I've been learning Rust on my new handheld gaming device/pc and think my setup is cool.

Also, I think learning new skills is a good use case for LLMs. I'm using Mistral (french privacy friendly ai company) to get guidance as I have questions, and the cost has been $0.67 so far this month.

This is very illuminating, all of my audio gear has resistors in them. Can't wait to tell my spouse and neighbors I've found the problem!

("On face value" might be doing a lot of work here. Not sure of his intentions. But in a best case scenario, I'm envisioning AI that is provided as a public utility/service, like the postal office, with people getting a say in its funding/training/deployment etc. No private companies profiting.)

Man, people are really jumping on this statement.

On face value at least, I think this is a very good place to start from with AI. What we have now is AI governed by oligarchs to serve themselves. If Canada effectively distributes that power to the people, that will be huge..nontrivial problem tho.

Ya I understand that. But similarly, it seems like if traffic is being captured that includes encrypted passwords, then those might be decrypted, no? Like, yea I get that historical messages being decrypted is bad, but you wouldn't want those same actors to take over your accounts as well?

Whoa, 33 months is not far off at all!

It seems like this should also probably be further motivation for people to setup 2FA on their accounts if they haven't already (ideally with a hardware key like yubikey or google's titan key when possible)? Or am I misreading the risk surface?

Grand Mountain Adventure: Wonderlands on Steam Grand Mountain Adventure: Wonderlands lets you explore freely and at your pace entire ski resorts and the surrounding mountains. Unlock ski lifts and reach new areas by competing in challenges, or enjoy the scenery while you find your own path down the mountains.

hell yea! i loved ssx3 so much.

i've started playing this game recently and it's cool - it reminds me of ssx3 due to the open-world-like setup and good mechanics that feel satisfying/realistic, but it's also a bit chiller and different pov. (it's on both steam and switch fwiw)

You might be right. But you don't need be on a big team to feel pressured by leadership to deliver faster. And I think a lot of people see the recent leadership changes at Bluesky along with feeling like Bluesky is pushing AI via Attie and are uneasy about it all. Users are expressing that.

AI is legit accelerating problems in large systems. Leadership at these companies prefers to blame workers for the problem they created.

I can also see how many folks feel that AI is empowering them and leveling the playing field and feel defensive of what they see as a democratizing utility.

Listen, I work with AI. I get that there's nuance and that AI isn't the source of all tech problems. But I've also seen first hand how companies are using AI as a tool to put pressure on tech workers, forcing them to deliver at unsustainable rates at the expense of oversight.

No, it's not.

Blaming everything on immigrants is punching down.

Blaming everything on AI is punching up.

AI is largely owned by the most powerful people/companies in the world. Without reasonable protections in place, people being hurt by it highlight/leverage the real problems it has created.

jeez that's infuriating! what's the recourse (if any) on your end? i didn't see your rant and am curious if there's anything that can be done about it in general.

My understanding of the risk surface is that certain companies/entities may be capturing encrypted data with the expectation of being able to decrypt once it's feasible, and that that will be very valuable (even if it isn't applied to everyone), so might as well start quantum-resistant encrypting?

this slaps and i didn't noticing the mouse dancing until it did that thing that made me smile way more than it should've. a++

it's ok to make and share stuff that is unpolished. point your phone at your handheld or tv and record your silly gameplay videos!

100%. part of it i think is that management often expects like 2x delivery speed, and engineers are using LLMs to help meet the new demands and just crossing their fingers

yea! i just went down a rabbit hole looking at their profile and seeing all these accts they interact with where if you look through their posts/replies and click the links in their profile, it becomes obv they're also ai. they're getting clever and a lot harder to detect at a glance..like wth

(fyi that's an ai bot making an argument for why LLMs are totally not a problem in any way in software engineering)

(i feel like i have to acknowledge that my own profile checks some of the boxes, and like, i swear i'm a real person, but wouldn't blame anyone for questioning...)

i honestly don't get what the point of these bots is. it feels like their purpose is just to get in the way of humans connecting online

their best lives and super productive in their creative endeavors.

god i feel so dumb. what a buncha dystopian bullshit. i'm sure some of them are real, but that a way higher percent than people realize are just AI.

@support.bsky.team please fix this truly insane problem you guys have!

and like, some of these AI accounts have their "thing" being *complaining about ai*. it's getting genuinely hard to detect and it's kind of terrifying tbh.

the trans thing is fucked up too. so many trans artists come across my feed, and I'm always amazed at how many trans people are living

and here's the thing - if you look through its profile, you'll see other accounts that it reposts and commonly interacts with. some of them have photo-like profile pics and will have a link to a soundcloud or youtube channel with ai-generated music, or to a blog/website with ai-generated nonsense

bluesky thread started by SpaceTessContinuum (@eigenvalue . bsky . social), in which it posts many quick, obviously LLM-generated responses in succession

profile for SpaceTessContinuum (@eigenvalue . bsky . social) Witch Queen of Seattle Transfem Hacker Send me a message!

check this out, this is pretty nuts...

after replying to a thread about coding, i got a flurry of responses from op in the span of a few mins, obviously LLM generated.

the acct (SpaceTessContinuum) says "Witch Queen of Seattle, Transfem Hacker"

this AI-posing-as-trans shit is gross and dangerous

No you're shifting the goal posts of what we're talking about.

I mean, ok but writing code and having a misunderstanding of some details is not really the same as vibe coding an application and having zero idea of how it works is it? Surely we can acknowledge that the less human accountability there is in development, the greater the risk of critical issues?

Wait is that true? Or is it just saying that all code is written on layers of abstractions that orchestrate an incredible sequence of physical events that no coder would be able to explain in detail? If so, is it not enough to be able to explain your code at the layer at which it was written?

Wow you worked on the art for Katana Zero? I just played it recently and loved the art so much. Really unique and a big part of what made the game special...great job!