Define unrestricted.
It's quite a stretch to read particular person, place, and thing as meaning individual directories on a computer or smartphone.
It's like restricting warrants to individual rooms in a home, a no-go, but in some circumstances I can see that making sense. Can't search the other duplex unit. Or you can search their phone, but you can't log into their bank app and download statements.
Is this some kind of media effort to claim there's really no difference between an LLM and a search engine?
Please actually ask someone who understands algorithms and data structures why this might not work out. The fact google have poisoned their search with advertising does not mean that's normal.
https://en.m.wikipedia.org/wik...
I have a dream for the Web [in which computers] become capable of analyzing all the data on the Web - the content, links, and transactions between people and computers. A "Semantic Web", which makes this possible, has yet to emerge, but when it does, the day-to-day mechanisms of trade, bureaucracy and our daily lives will be handled by machines talking to machines. The "intelligent agents" people have touted for ages will finally materialize.
Tim Berners-Lee, 1999
We're finally beginning to realize this goal, using LLMs. This talk about what a true search engine is, it's funny. No, they're not search engines powered by ad-network heuristics or link popularity, and they're not indexes powered by keyword matching. I guess you could call it semantic search because the web didn't change to accommodate machines, machines caught up to us, or our junk we put on the web anyways.
/. isn't what it used to be, but more importantly, the world in which
Well, that and no CmdrTaco.
Exactly, and here's Newsmax of all things calling out the administration over a big part of it.
https://www.newsmax.com/newsfr...
They preemptively fired their own lawyers because they want yes-men for legal advice, or more accurately they don't want or care about anyone's opinions on the law and constitution.
JFC, fuck off. How about we scrap your insane idea and invest in other countries. You'd rather go to war with China than invest anywhere else with cheap labor? See that doesn't add up.
The only way you know out of a bad situation is to do something worse and say it had to be done. We all know people like that around us in every stage of life, they're morons and can't be trusted with anything.
But he stopped all the wars! That one with Iran, there was Eastasia, there's no more fighting in North Korea, windmills are no longer a threat, Christmas and drugs are safe, and the big one, Godzilla will never be back.
Hang on, I need to get my boss a towel.
The hallucination problem _cannot_ be fixed. It is a fundamental part of the mathematical model. Getting it fixed is about as possible as making water not wet under standard conditions.
I can't fix your hallucination problems either, but through training, repetition and reinforcement, we can overcome your fear of the dark even if the notion that there's a there there never goes away.
We could reduce this to no model of reality can be perfect. I don't need you to be perfect to be useful.
I, for one, have a lot of two-digit numbers to factor, so I'm waiting impatiently for quantum supremacy.
Slashdot's front page led users astray; reports duplicate stories are new.
When TFS or the headline are misleading, you head to the comments to complain about it. Maybe a few people read TFA. Engagement goes up anyways.
When the Google AI summary is wrong you close the browser tab because you got what you came for, as far as you know. Even if you knew it was wrong, how do you let others know?
I don't want to in any way defend engagement bait headlines, but at least their purpose on a place like this or a call-in show or whatever is to argue about it, not to send you on your way totally misinformed. Intentional or not, confidently wrong is dangerous. Gemini proper should come with the same warnings like any other LLM, and here too, this place is for entertainment not news, etc, etc.. but the auto Gemini summary thing though. I've been comparing LLMs to power tools and between a user pressing a button and the motor engaging, we should maybe put a little more thought into what's happening there sometimes.
You can't get away with a good rape joke anymore, but what does that have to do with being funny... or making movies?
You're trying to draw a line from what's funny through what's acceptable to what makes money and ended up someplace absurd like you couldn't get away with making South Park today. Well, you couldn't do it and make lots of money. Well, m'kay boomer.
How do you maintain all the translated copies, this isn't rocket science, dumb dumb. If there are enough interested editors with fluency in a given language go ahead and submit it, otherwise stick to automatic translation.
Your hypothetical "what if I want to research Nazis" is transparently bullshit. For example if Timothy McVeigh wrote his own page on Polish Wikipedia, how the fuck does it help Polish readers who now see THAT instead of the American consensus on that shitbird.
The self promotion isn't even the worst problem, this is something you do when you have something to hide - because the GOAL is nobody editing your bullshit in all languages.
Exactly. Ever see the front page of Slashdot without an adblocker?
When CmdrTaco introduced ads, they were at least RELEVANT to the site... ThinkGeek, etc... Now we get Temu shit.
Right now I see Chuck Norris next to a peeled banana being sprinkled with cinnamon apparently, text says three foods to always avoid.
Maybe ad networks are the problem... not advertising. There's a huge difference between ThinkGeek banner ads and something that never in a million years anyone would say I need Chuck Norris next to a banana right across the top half of my site selling what is 100% a scam.
I'll take oxymorons for $100, Alex.
Maybe I'm biased towards "legacy media" but I never flipped through a Popular Mechanics and thought the ads sucked. There are still plenty of whatever that was, curated ads, collaborations with advertisers, etc but online, it's usually mixed with ad network garbage.
Look at this. Nice ad for a Swiss Army knife integrated with the site at the top, that's tasteful, very on brand, but there's the random shit with little green triangles, Google's ad network below.
https://www.popularmechanics.c...
Then there's the very bad end that's all taboola garbage some sites have, Chuck Norris super imposed with a rotten banana and some thing to fix whatever problem that is. If you don't try to discern the difference you deserve the garbage you get.
Or a
At one point programmers used to brag that they could learn any programming language in a day or a week.
Now apparently there's a substantial portion of the programming population that can't learn any language without AI assist.
I think you're misremembering.
https://www.norvig.com/21-days...
Or the 1998 version
https://web.archive.org/web/19...
It's been understood for a very long time that learning a language is the easy part, especially after your first. Mastering its libraries, frameworks, its isms, that takes considerably more time, Peter's Ten Years suggestion.
That deep end of the pool is where AI really is invaluable for.
Disclosure: OFC I used AI to find notable people in tech with blogs from the dotcom era supporting my point. How the fuck do you Google that?? It's also what I remember from the time. You can and should learn new languages because it teaches you to look at the same things differently. But we all knew what the deep end consisted of. Whole books dedicated to one particular end of a system, like Java Swing for example, or OpenGL. AI helps you carve through all that.
May all your PUSHes be POPped.
