Interviewed at two big, well-known tech companies. Had done a lot of mobile dev work at the time, but really wanted to switch to connected hardware and told the recruiters.

Showed up for the first on-site interview. Guy walks in. Explains the actual first interviewer couldn’t make it so he was a last-minute stand-in. Goes: “So, it says here you are intererested in mobile. That’s good. My team is looking for someone like that.”

I explained it was actually the other way round. What proceeded was an awkward hour of bullshit questions about train schedulers and sorting algorithms. Repeat five times that day. Every. Single. One.

Second company a few weeks later. Same thing. Except this time, 2/3 of the way through, a manager in HW group walks in. Grouses why he was asked to talk to someone, checks notes, about mobile. We had the greatest conversation after I set him straight. He wanted me to come back and do another loop just with his group. Except a week later, they announced a hiring freeze and I never heard back.

In retrospect, it was a good thing. I would not have been a good fit.

Now only have to wait for:

• pyenv release
• pycharm update (including terminal)
• 3rd party libraries

to catch up…

Good thing they stopped emptying train toilets on the tracks.

Tried it with ChatGPT 4o with a different title/author. Said it couldn’t find it. That it might be a new release or lesser-known title. Also with a fake title and a real author. Again, said it didn’t exist.

They’re definitely improving on the hallucination front.

The repository included the private portion of the platform key in encrypted form. The encrypted file, however, was protected by a four-character password, a decision that made it trivial for Binarly, and anyone else with even a passing curiosity, to crack the passcode and retrieve the corresponding plain text.

It’s like installing a top-of-the-line alarm system for your house with camera, motion detector, alarm, and immobilizing gas, then leaving the unlock password on a PostIt under the welcome mat.

I intentionally kept historical imports out, since Reddit is blocking APIs under the guise of limiting AI scraping.

My main point was setting up an easier way for low-tech mods to set up a parallel community, then nudge users to move over.

Could there be a one-click way to automatically ‘import’ a Reddit subreddit over to a Lemmy community? Meaning, create it, import the sidebars, welcomes, rules, graphics, etc. so it looks familiar to regular users. If not, at least a step-by-step tutorial on how mods could do it.

Another option would be to provide something like a crossposting Chrome or Firefox extension that lets people simultaneously post content to both Reddit and Lemmy. Give them a smooth transition path.

Lastly, the Bluesky concept of ‘pluggable algorithms’ is one way to make it so users can choose whatever sort works best for their interests.

No shit. People have known about the perils of feeding simulator output back in as input for eons. The variance drops off so you end up with zero new insights and a gradual worsening due to entropy.

Threads for daily games of favorite MLB teams. Where people posted realtime takes during the game.

Thank you. Looks like I’m not alone and people are doing more detailed testing. I’ll just wait till the dust settles.

I tried downloading and running the 405B locally through LM-Studio. Got an error message saying invalid tokenizer. Then tried it with ollama. That didn’t work either. Going to try the 70B tomorrow.

Not sure it’s possible to run the larger ones on a Mac laptop.

Every time there is a power shortage, they tell residential customers they are being wastrels and should turn their lights off…

Same for water.

AWS bread and butter is EC2, S3, and Lambda.

The reason AWS is focusing so much on Gen-AI is because they’re in the shovel-and-pick business during the gold rush. This guy’s beef should be with the over-excited gold speculators, not the general store purveyors of denim and panning equipment.

Tried running it on a Mac via LM-Studio. Wouldn’t run. Said llama.c needed updating.

They all add donors to lists and send them solicitations based on the premise that a former donor doesn’t need as much convincing for their cause. It’s fairly easy to unsubscribe as soon as you get the first message, but by then your email may have already entered a partnership pool, so you’ll start getting similar solicitations.

Best way around it is to use an email anonymizing service, or one way phone/text numbers. Or you can treat it like a game and donate under username+charity@mailservice.com, then watch how many more messages you get to that specific address.

To answer your main question, eff.org and msf.org are pretty good charities. You can also check https://www.charitynavigator.org/.

If it’s purely political orgs, local groups are less likely to have aggressive fundraising arms.

If you were to design a government system from scrstch, knowing what you see now, where would you place a Supreme Court?

ESP-32 based voice controller, with LLM support: https://www.home-assistant.io/voice_control/s3_box_voice_assistant/

• Monty Python and the Holy Grail
• The Big Lebowski
• Inception

Edit: Can’t believe I left these out:

• High Fidelity
• Princess Bride

It would be amazing if parts of it could be animated. Maybe multiple layers of canvases (say, 5 frames, shown over a second). Each with their own images, which could be viewed as a flipbook.

Instead of going for a larger canvas, go for more layers.

Just a thought.

This has been one of the most wholesome and heartwarming things I’ve seen online. People behaving well and working together. Kudos to the organizers and all the participants.