The watermark would likely be comprised of a few different methods to embed marker pixel sets that would be difficult/impossible to see in addition to ones that are visible. Think printed currency. I’m not saying there won’t be an arms race to circumvent it like drm, or bad actors who counterfeit it, but the work should be done to try to ensure some semblance of reliability in important distributed content.
Someone being disrespectful isn’t even close the being the government’s business. You can be as disrespectful as you want, burn a Koran every day if you please, and the government should not (or be able to) do anything about it.
Most religions try to impede the rights even of nonmembers of that religion. That is disrespectful as well. Burn any book or other goods you want, if you want to protest any of my convictions. TuX-puppets, books about string theory, the FSM, my national flag, I don’t give a shit as long as you own those things and you don’t physically harm others.
So is calling for death to dissenters and unbelievers. Actual murder of them is morally abhorrent. That god and that prophet are no more special than the thousands of others remembered and forgotten.
LLMs choose words based on probabilities, i.e. given the word “blue”, it will have a list of words and probabilities that those words should follow “blue”. So “sky” would be a high probability, “car” might also be quite high, as well as a long list of other words. The LLM chooses the words not by selecting whatever has the highest probability, but with a degree of randomness. This has been found to make the text sound more natural.
To watermark, you essentially make this randomness happen in a predefined way, at least for cases where many different words could fit. So (to use a flawed example), you might make it so that “blue” is followed by “car” rather than “sky”. You do this throughout the text, and in a way that doesn’t affect the meaning of the text. It is then possible to write a simple algorithm to detect whether this text was written by an AI, because of the probability of different words appearing in particular sequences. Because its spread throughout the text, it’s quite difficult to remove the watermark completely (although not impossible).
Self driving cars are a bad idea. There, I said it. It’s solving the wrong problem with technology that is nowhere near ready. The world is simply too dynamic and the “edge cases” matter.
Better safety features, however, will be a great side effect of this research though.
The thing is it’s been “real close” for ~10 years. But they’ve solved the “easy” problems. In development you spend 90% of your time on 10% of the problems. This is why fusion power has been “real close” for 30 years as well. Those remaining problems are the hard ones.
I genuinely don’t know why Tesla doesn’t just focus on the safety aspects of their tech rather than the “self-driving” BS. Having a car that will stop for me if the car in front slams their brakes on unexpectedly is a great thing. That’s a lot of accidents avoided.
Then take a train or a bus. I doubt we’ll ever get to the point where self-driving cars are reliable enough that you could legally sleep while it is in operation. Especially not in our lifetime.
Did I say autopilot was safe? I said the Tesla Model Y was the safest car - you know, by test crash ratings.
I hate Musk as much as the next guy, but people can’t separate that ass from the actual good things produced by some of the companies he owns.
FSD is fine if you still drive the car. People don’t do that, so it’s actually dangerous in the real world. It should be marketed VERY differently.
If you get into an accident, being in a Tesla is actually safer than a lot of cars because the car itself is very safe.
Of course everyone hates Musk, so anything related to Musk has to be down voted. Any opinion other than Space-X and Tesla is as shit as Twitter and Musk himself is wrong.
It sounds like your saying that self driving cars moving to the consumer market too soon is more of the problem. Am I understanding that correctly? If so, I agree. I think the tech is more like 10-15 years out still at least. There may be other smaller applications sooner but the continued improvement of safety features can be done now.
I’m not even sure 10-15 years is good. They’ve been 10-15 years for 10-15 years already. I’m personally at the point where they’re square in the “flying cars” category (which is coming “real soon now”!).
Detecting things is easy. Finding the road and following it is easy. Stopping the car when an obstruction is in the way is easy. What’s not easy are the 100,000 things that the developers haven’t thought of that happen in a real-world dynamic environment. And it’s a situation where lives are at stake so you need to get those right.
And then there is the issue that we already have “self-driving cars” in the form of light rail, busses, taxis, etc.
I think the combination of human driver with “AI Assist” for cruise control, avoiding obstacles, and other things is likely the way to go for cars for some time.
There cannot be topics forbidden to be made fun of in a free society. If we cannot make fun of Muslims and Islam in any way without causing religious riots and violence then current form Islam has no place in a free society. The difference between Islam and the other monotheistic abrahamic religions is that the older two of the three have had enough time to go through cycles of reformation and ideological modernization, making them somewhat compatible with a modern pluralist society. Islam still insists a bunch of fables from when the book was written are literally true and valid today, and not just apocryphal advice on how to be a decent person.
And before someone brings up the inevitable American evangelicals as comparison, they are equally incompatible. Luckily that scourge is largely confined to the United States and not my problem.
Topics can be forbidden while still allowing for a free society. It’s called the paradox of tolerance and is why Germany bans the free speech of Nazis. The thing about satire of Islam is that the events that get a big reaction like this are not “making fun of Muslims and Islam” but flat-out hate speech. Look at real political cartoonists like Eli Valley and Stan Kelly to see that you can joke about and even be kind of mean to people without calling them pedophiles or burning sacred books. You say that Islam hasn’t had time to become compatible with modern society. What a ridiculous statement contradicted by your own exception. The rise of American Christian fascism as well as the arch-conservative Christians in modern-day Uganda prove that time has nothing to do with how strict a religion is. This is further contradicted by Islamic society being much more progressive in the Medieval era compared to Christian society despite being younger. The fact is, you just hate Muslims.
Thank you for enlightening me to my inner workings! It is always a revelation to have an internet armchair psychologist give you a diagnosis.
For that matter, I am german myself and thus have a “passing” familiarity with the legal situation surrounding nazism. I am specifically referring to Christianity in the west, with an exception to America due to their own regression in that regard. You are of course right that certain re emerging right wing parties are sometimes tied to conservative Christianity like in Poland, but that doesn’t somehow negate the danger of allowing a new and radically violent religion to take hold and carve out space in a free society.
Also, your paradox of tolerance is not actually a paradox, people just like to parrot that to not be made to change something.
Tolerance is a social contract that says I accept you to live and be free as you are or want to be in this society, as long as you treat me with the same respect. Islam does not subscribe to this notion, there is no room for acceptance of LGBTQ people, the equal rights and opportunities of women, not even for non-Muslim beliefs (infidels) or atheists (apostates).
Which in turn means Islam is not covered by the social contract of tolerance.
Cranston accepted Shapps was wrong to describe Naumenko as a friend of Putin, but said it was “excusable political hyperbole”.
Um… I wouldn’t dismiss it as “excusable”. The fact that he’s not friends with Putin is the whole point. It feels really weird to comment in favor of a Russian oligarch, but come on. Justice is more important, and this doesn’t seem fair.
There is such thing as collective responsibility. He lives in Russia, pays taxes in Russia, they go in support of war. Yes, it is unfair to him, but unfairness is the fact that he was born in Russia, he is Russian citizen, and it is Russian government actions that triggered everything.
Regardless, he’s saying this stupid shit for more than 10 years now. Self driving cannot be solved without an artificial general intelligence because there has to be an understanding of what other people are going to be doing
If there was dedicated lanes/infrastructure it might be possible but makes more sense for cities to improve public transportation. A bus/train is a big fancy car powered by a general intelligence.
Eh. You can probably solve it with a good enough artificial narrow intelligence. Or/and dedicated infrastructure, inter-car communication protocols, etc. The issue is it's solving the wrong problem altogether.
Years ago (maybe still) Microsoft had a research facility for self-driving infrastructure. Instead of putting all the recognition and awareness in the car itself, a lot of it was offloaded the mini city they built. Streets and stop signs with embedded RFID, etc.
This, of course, doesn’t stop pedestrians from dying. But I thought it was a cool approach to the problem to “update the world” instead of trying to make a product that navigates our unmodernized infrastructure
Maybe, though trams only work in town. I couldn’t go see my family with a tram but I could put my self-driving city car in manual and take it out past the cornfields.
I think a lot of things have to change outside of major cities for public transportation to really take off as a concept here. There is SO much “empty” space in the US, it’s hard to imagine getting infrastructure out there that mainly only benefits a handful of people
I think they used to include RADAR in their cars, which is probably better for handling weather conditions that would interfere with light based systems (fog, snow, rain, etc.). They took it out, with Musk claiming they could do FSD with just cameras. Probably it was about cost or supply, and I think they decided to add it back recently.
reuters.com
Active