Mark T. Tomczak

Yep. This one is a maddening head scratcher, because I have no idea how we would pursue it deeper without trying it, I definitely am not volunteering to be first in line to chase this particular thought experiment, and I don't even know how we would validate the word of the subject on their personal experience.
(... That last one may not even be a major issue. The solipsists will remind us that we can't actually pull that trick off with the other meat brains around us. ;) )

2024-06-20 18:12:09

- reply

Ah, you've just inspired me to bother to look at my robots.txt file.
... looks like Mastodon node doesn't disallow the Google crawlers by default, so I assume it's not unexpected your content would show up in Google's search engine? Though one thing I've never stopped to wonder: if you *do* disallow Google's spider, but your posts are imaged to someone else's node because someone on their node follows you, can the spider pick up your posts from there? This post on your server also has [https://mastodon.fixermark.com/@[email protected]/112649798768708524](A URL on my server), so I'm *pretty* sure the spider could see your post via my node even if you disallow it on your node.

2024-06-19 11:04:26

- reply

Those are good features. I'm trying to wrap my head around how one would implement them in a Fediverse context... You could have one node say "I won't track replies from other servers / advertise them to peer nodes," but how to force every other node to comply?
This smells much easier to solve in a server monoculture.

2024-06-17 17:29:20

- reply

Surely we can trust this feature in the hands of YouTube commenters, the most responsible and polite demographic of commenters on the Internet, who were never almost single-handedly responsible for killing another social media experiment when it was merged with the YouTube comment stream.

2024-06-14 18:57:51

- reply

I think the next logical step is to apply this question to self-driving vehicles.
... though there's likely already precedent there: who is responsible when a malfunctioning autopilot crashes a plane?

2024-06-12 23:36:44

Having one of those "Rerouting my DNS through Google's free servers because my ISP's machines are choking today" kind of days.
It's... Hard to care that it feeds useful data to Google when it's so damn fast and never goes down. Be nice if the rest of the web were this reliable.

2024-06-11 17:58:18

- reply

TBH? I don't know.
Democrats far more than Republicans tend to seem to cling to previous-generation ideals of how government is supposed to work, even in the face of obvious evidence that those rules no longer apply. To my mind, the question of whether he'd be willing to let his son hang out in jail a few years longer to try and "heal the nation" is a coin toss right now.
(Were it me, I'd probably pardon him and say "If you don't like it, take it up with my advisors, Left Middle Finger and Right Middle Finger," but this is one item in the *large* pile of Reasons Nobody Should Ever Elect Me President. 😉 )

2024-06-11 17:53:20

- reply

As is within his power, though of course if he does so it will raise *huge* questions of nepotism.
But honestly, this is a story of the system working and while I don't doubt it's heavily politicized, I'm appreciating watching some rule-of-law happen.

2024-06-10 18:27:14

- reply

It was kind of an eye opener for me the day someone noted that the government could, technically, take all the money collected in taxes, set it on the Washington Mall, and set it on fire... Then print that exact same number of bills the next day at the mint and (apart from the ridiculous carbon footprint) it would *change nothing.* And they can also choose to print more. Or less.
Taxes in a modern fiat system just don't work like taxes used to. We don't teach this well at all.

2024-06-05 20:24:38

- reply

Interesting! I haven't seen maps disappear from my search results yet; is this an EU-only thing?

2024-06-05 20:16:55

- reply

What's the context?

2024-06-01 20:51:23

- reply

Best of luck!

2024-05-22 21:51:23

- reply

I suspect that without S230, Mastodon is not practical for the US.
If every site owner were legally liable for everything said from their node (or, for that matter, everything *transited through their node,*) uh oh!

2024-05-21 23:31:39

We all know that this new Windows #recall feature is a concern, but we also all know it's going to be hugely popular right?
Because the flip-side of the coin for "domestic abuser surveillance" is *parental surveillance.* Parents trying to figure out how to raise their kids safely in the digital age will *gobble up* the ability to check on what their young ones are doing online when they're not around.

2024-05-16 21:57:33

- reply

Sorry, I think you're operating at the tip of a chain of reasoning that I don't see right now.
That's okay. I'll think on it and maybe I'll come around to your view. Your passion on this suggests to me you're perceiving a danger I'm not seeing here.

2024-05-16 21:45:50

- reply

Sure, but accusing Google of pulling some kind of shenanigans because they *added* a UI feature to focus your query (over a default that they believe is most useful to the most people) seems... Odd?
Google's mission statement is "To organize the world's information and make it universally accessible and useful." The mission statement *has never been* "To send people to websites." Websites were always the means because that's where the information was... Is it anymore? If Google can give you your result faster without navigating you away, it's in their mission statement to do so.

2024-05-16 21:44:01

- reply

But what does that mean? The algorithm *itself* (guessing from usage patterns what humans probably mean when they give a query) is "AI crap." Onebox has been around forever and it's "AI crap." From Google's standpoint, this is just an evolution on a tech stack they've been building for decades now.
I agree with you though that it'd be nice-to-have to be able to sticky that "Web" option---really, any of those options. Or get to them by a keyword ("in:flights") for much more convenience on the input line.

2024-05-16 21:36:08

- reply

Isn't this a good thing? People who want to bypass the generative content can just go to the "Web" tab now. Granted, I'd like a way to shortcut into it like images.google.com taking me straight into Image search, but I imagine writing a Chrome extension to wallpaper that shouldn't be too hard if clicking the "web" button becomes a burden.

2024-05-15 22:00:35

- reply

Figuring out code is tainted by use of copyrighted code from another source is as straightforward as string-matching, maybe some fuzzy matching.
How would one identify code generated with the assistance of an LLM if the contributor doesn't admit to doing that?

2024-05-14 20:28:49

- reply

I worked with a colleague at Google who's byphrase was "Don't optimize for the unusual case." Every time I hear about someone losing their account permanently and being lost in the labyrinthine halls of the recovery process that assumes a particular type of spherical point-like user, I think about the fact that he still works there and his philosophy is closer to the centroid of Google's philosophy on software than mine.
... anyway, a friend's parent got his Yahoo account compromised a little while ago, and he was able to reclaim it with one phone call. Someone at Yahoo checked the recent access logs, confirmed that it was obvious that someone had radically changed the access pattern of the account immediately after a password change, and *just reset the password back.* No muss no fuss.
So now I tell people to get a Yahoo account if they can.

2024-04-19 18:43:08

#ai #openai
"AI is stealing from artists" is an argument that falls flat. This is not to say AI doesn't interfere with a traditional artist's ability to turn their labor into income; just that "stealing" is probably the wrong term.
Google (and other search engines) have been collecting, indexing, and making searchable the Internet for decades. They do that by exhaustively crawling and copying data from other people's websites and transforming it into a novel representation. Modulo some dissent, nobody really considers that crawl itself "stealing," nor the building of hte index.
But notably... Where people *do* get bent out of shape is when Google crawls the Internet into OneBox and provides zero-clickthrough factual answers to questions. *This* has generated some heat. But the heat is not grounded in a philosophical "theft," because it's not about how the data was harvested or indexed, but how it was *used.* The search index *directs traffic to other sites,* the OneBox answers questions *without redirecting user attention to the source of the information.*
"Theft" is a bad term for this. It breaks down because we do not have a common-sense understanding of "theft" being "how the acquired property is used;" we have a common-sense understanding of it in the taking. OpenAI wasn't "stealing" from artists when it was crawling and indexing the world of images; it wasn't "stealing" when it was labeling data or building bridges to existing label caches; it wasn't "stealing" when applying training data to them or comprehending the training data. If they'd made a big art search engine, I doubt we'd be having a "stealing from artists" conversation. But use that dataset to synthesize novel images and *uh-oh!*
So there's probably a better term for what OpenAI does than "theft," but I don't have it. And I don't think some people want to find it because moving the conversation to the use of the data opens the question "Was something actually taken from the source artists *that they deserve to have*", and that's, I think, an open question, one that many try to short-cut by falling back to an idea of "theft."