More network state nonsense is afoot: https://frontiervalley.com/ https://www.thenerdreich.com/startup-seeks-trump-ai-emergency-for-california-tech-city/

They (who?) have publicly drafted an executive order because they want to take over the Alameda Naval Air Station (a superfund site).

Edit: Per the twitter account the weirdo behind this is James Ingallinera.

Every time I see one of these AGI meltdowns I’m reminded of this anime / manga:

I’m sure it’s a complete coincidence that the people who think it’s impossible to program without StackOverflow are the same people trying to shame us for not being LLM enjoyers.

Edit: But also - why do AI scrapers request pages that show differences between versions of wiki pages (or perform other similarly complex requests)? What’s the point of that anyway?

This is just naive web crawling: Crawl a page, extract all the links, then crawl all the links and repeat.

Any crawler that doesn’t know what their doing and doesn’t respect robots but wants to crawl an entire domain will end up following these sorts of links naturally. It has no sense that the requests are “complex”, just that it’s fetching a URL with a few more query parameters than it started at.

The article even alludes to how to take advantage of this with it’s “trap the bots in a maze of fake pages” suggestion. Even crawlers that know what they’re doing will sometimes struggle with infinite URL spaces.

Hell yeah!

Seems there’s a lot of buzz about pride month this year. I’ve been to one pride parade as a teen and have approximately zero LGBT fashion items, but solidarity and visibility seems more important in recent years. I should find a necklace with trans flag colors or something.

LLMs: now as effective as enumerating use-after-frees as grep "free" source.cc.

Holy hell all the examples I found made me seasick. I am apparently physically incapable of watching veo3 videos.

The extension I ended up using for Firefox was Straight to the Web. Auditing the source code I saw that it looked for certain Google URL patterns being navigated to and re-wrote them.

I just tested with DuckDuckGo’s “!g” feature and it seemed to work, but I don’t use DDG so I don’t know if there’s anything I’m missing.

Yeah I’ll probably have a big tax bill if I ever renounce citizenship. I haven’t thought about it too much yet since it’s still my only citizenship, and I have a lot of friends and family in the USA. Like being a visitor might be fine in normal times, but I wouldn’t want to rely on it in an emergency today given how visitors are being treated lately.

'Till now I was always able to just do financial planning myself, but I really should hire a professional.

An internal transfer at my job actually. At least for now they need me so helped set that up, though I’m pretty worried on if that will last long enough for permanent residency or not.

I’d be a little nervous on a job seeker’s visa before knowing the language. It is really hard to find a job as a programmer in Europe without living there or being a citizen; because of language barriers, the labor market test, and the difficulty in getting a company to sponsor your visa. I didn’t send out that many job applications but so far my response rate is zero.

Probably if I couldn’t do a transfer I’d have ended up on an investment visa or study visa somewhere; though maybe I could have found a job in Japan since I can read intermediate Japanese.

I expect learning German to the B1 level will open up a lot of doors, so that’s my main goal for the next few years.

Nah it’s not too bad the IRS guide is only 40 pages! ^{somebody save me}

• All US citizens get to file US taxes every year regardless of if they have any US sourced income
• Foreign income is also taxed (but see next two points)
• The first 126k of foreign income earned while living abroad is excluded from taxation (Foreign Earned Income Exclusion)
• Income that went to paying foreign taxes is also not taxed (Foreign Tax Credit)
• Banks hate opening accounts for US citizens since we’re subject to FATCA filing requirements and thus generate extra paperwork
• Also certain foreign mutual funds are taxed heavily (PFICs), requiring care in planning investments.
• There are a bunch of tax treaties with different countries, which may influence the exact details.
• If you do have deferred compensation that was granted in a state but that was vested or exercised while a non-resident of that state you may also have to file state taxes (e.g. FTB Publication 1004 for California)

I haven’t run through this in practice yet and I will probably give up and hire a professional.

OK completely off topic but update on my USA angst from earlier this year: I’m heckin’ moving to Switzerland next month holy hell.

Back on topic: Duolingo continues to circle the drain. I kind of hate that I’m linking to this because it’s exactly what that marketing-run company wants; but they posted these two videos to TikTok in response to the AI backlash: https://www.tiktok.com/@duolingo/video/7506578962697456939?lang=en https://www.tiktok.com/@duolingo/video/7507337734520868142?lang=en

I uh… I don’t think it’s going to change anyone’s minds. Half the comments on the videos go something like:

EVERYONE LISTEN UP!!! 🚨 - starting from today, we are gonna start ignoring duolingo. We will not like the video it posts, or view it. - BASICALLY WE WILL IGNORE DUO!!💔 💔 ON EVERYBODY SOUL WE IGNORING DUO!! 💔 (copy this and share this to every duo related video)

I had been using a CSS rule to hide the AI overviews on Google.

Because of this news I just installed a Firefox extension to force Google search results to use the “web” tab (which presumably skips generating them entirely).

[4:00] … whose products [I] actively think are at best valueless and at worst harmful

Wow what a scathing critique of worldcoin! Calling it possibly harmless. Clearly even when making the apology he didn’t really get why we all hate it.

Here’s a video of a Tesla vehicle taking the saying “move fast and break things” to heart.

Ah yes the typical workflow for LLM generated changes:

1. LLM produces nonsense at the behest of employee A.
2. Employee B leaves a bunch of edits and suggestions to hammer it into something that’s sloppy but almost kind of makes sense. A soul-sucking error prone process that takes twice as long as just writing the dang code.
3. Code submitted!
4. Employee A gets promoted.

Also the fact that this isn’t integrated with tests shows how rushed the implementation was. Not even LLM optimists should want code changes that don’t compile or that break tests.

Ugh. So terrible. Tech’s obsession with “scaling” is one of the worst things about tech.

Yeah that jumped out to me. Like human teaching has scaled fine to billions of people. It certainly has a better track record than Duolingo which provides meh study material and leads to ahem mixed learning outcomes despite being around for over a decade.

Of course there’s the subtext of “but also we’ll be able to put all those obsolete teachers out of business and make tons of money!”

Aaaarrgh. Tech’s obsession with A/B testing is another one of the worst things about tech.

Being in tech I definitely see misuse of A/B testing sometimes. Sometimes a team will ignore common sense entirely but come up with metrics that measure something irrelevant. The metrics are, intentionally or not, gamed to tell them what they want to hear. They then run the (useless) numbers and use that to justify why their change was good, even in the face of intense user backlash.

One particular example that just came to mind: someone made a bad change, and lots of people complained. Eventually the complaints started to peter out. Then they claimed “see! people just had to get used to it!” (versus the rather more obvious possibility that nobody bothered to complain more than once).

Duolingo CEO says AI is a better teacher than humans—but schools will still exist ‘because you still need childcare’

It’s wild for the CEO of an edutainment company to have this much disdain for for teachers.

can’t have AI bro coworkers if you’re unemployed :P

I’d certainly feel less conflicted yelling about AI if I didn’t work for a big tech company that’s gaga for AI. I almost wrote out a long angsty reply but I don’t want to give up too much personal details in a single comment.

I guess I ended up as a boiled frog. If I knew how much AI nonsense I’d be incidentally exposed to over the last year I would have quit a year ago. And yet currently I don’t quit for complicated reasons. I’m not that far from the breaking point, but I’m going to try to hang in for a few more years.

But yeah, I’m pretty uncomfortable working for a company that has also veered closer to allying with techo-fascism in recent years; and I am taking psychic damage.

Urgh over the past month I have seen more and more people on social media using chat-gpt to write stuff for them, or to check facts, and getting defensive instead of embarrassed about it.

Maybe this is a bit old woman yells at cloud – but I’d lie if I said I wasn’t worried about language proficiency atrophying in the population (and leading to me having to read slop all the time)