My perspective on this is that it’s fine to not like things and bounce off them. I also bounced off NMS, no big deal. It just isn’t for me and that’s fine.
All I know is that they worked hard on this game and found plenty of people who enjoy playing it a lot and that’s that. Doesn’t mean they are wrong, doesn’t mean I’m wrong in bouncing off. I do think it is interesting to explore why I bounced off and others might not – I can identify with many of the negative comments here, especially the 10,000 bowls of oatmeal one regarding the procedural generation which probably was the one thing I was most fascinated by when I decided to install the game – but there are so many other systems and elements in this game that if you more into those I can easily see why the game works much better for you.
Your rant seems confused from an art history point of view.
I think we should talk about the Bauhaus and where they were coming from. Industrial mass production allowed for cheaply mass produced ornaments during Historicism. To me it is easily understandable why the Bauhaus wanted to find a different approach to how to deal with industrialization and how to incorporate it into architecture, going away from mere imitation and copying. I think it’s extremely fascinating that those ideas took root in the Soviet Union (creating utilitarian living spaces on a mass scale) and the USA (office buildings).
Many contemporary artists, however, have very little or nothing to do with Bauhaus and its ideas at all. There is just no link connecting them … and while the Soviet Union did employ industrialized mass production to build living spaces without ornaments its art was more often than not Socialist Realism, which is a different beast altogether.
But I do recommend you read Asylum for the Homeless by Adorno because I think you might like it: “Actually one can no longer dwell any longer. The traditional dwellings, in which we grew up, have taken on the aspect of something unbearable: every mark of comfort therein is paid for with the betrayal of cognition; every trace of security, with the stuffy community of interest of the family. The newly functionalized ones, constructed as a tabula rasa, are cases made by technical experts for philistines, or factory sites which have strayed into the sphere of consumption, without any relation to the dweller: they slap the longing for an independent existence, which anyway no longer exists, in the face.”
You criticise the previous poster for seeing Communism everywhere and then recommend them a book by Adorno, a dyed in the wool red. That will confuse 'em.
You might be able to hide like a hundred drones in your backyard. Enough to kill a couple hundred people, maybe.
You cannot and will never be able to hide millions of drones in your backyard. That’s the relevant difference here. And that is not a mere quantitative difference, that is a qualitative difference that fundamentally changes the situation.
This is an incredibly dumb scenario that belongs in a cheap sci-fi novel because it hand waves away human motivations (Who should do this and to what end? What is there to achieve with something like this that cannot be achieved some other more conventional way?) and feasibility both in terms of actually getting the hardware ready and completely ignoring counter measures.
Which is to say: there are significant downsides to such an approach to the extent that this is fit for a spooky story or an abysmal horror flick but not compatible with actual reality.
Cheap drones do change a lot about how wars are fought, sure. They are in that sense extremely relevant. But that does not mean that your scenario or similar scenarios like it are remotely realistic.
Put another way: Cheap drones are not nuclear weapons (which actually did lead to a earth shattering change in global power dynamics, though I would argue even their transformative hasn’t been absolute).
Cheap drones are more like machine guns, tanks or airplanes. “Mere” new technology. Obviously hugely important and impactful, not earth shattering.
(As for whether AI might be earth shattering, I do not know nor do I even want to get into that question here.)
It literally re-weights the output tokens from what the LLM would otherwise have chosen. It _has_ to. It can't be positive, because then that's not watermarking, it's a better LLM.
To add nuance, that article does say one of the two versions does reduce text quality (and the other is worse at detectability):
> SynthID-Text can be configured to be non-distortionary (preserving text quality) or distortionary (improving watermark detectability at the cost of text quality).
What if the next token represents a wrong or low-quality answer, but would have only been picked 10% of the time, but now it's picked 20% of the time? Doesn't that obviously decrease the model quality, even though "it might have picked that token anyway"?
It would be picked 10% of the time with watermarking.
The randomness properties of the PRNG will be very similar to other random number generators, it is just chosen to be vulnerable to a particular cryptanalytic attack (that requires a private key known only to anthropic). I think of it like the Dual_EC_DRGB generator rather than a biased coin.
It is absolutely possible that it would not continue to be picked 10% of the time with a given fixed watermark key. The implementation literally labels tokens using a keyed hash and then modifies their scores. The entire point of the watermarking system is to bias certain tokens against others, and - as you would expect - this reportedly results in a reduced response diversity.
What if the token represents a high-quality answer, but would have only been picked 80% of the time, but now it's picked 90% of the time?
I'm not entirely sure (haven't read the original synthID proposal), but I believe that the re-weighing is set to make both your scenarios and mine equally likely, averaging out to net Zero effect on quality.
Unless you’re at 0 temperature, there is no single token it would have chosen. It’s always picking one of multiple randomly according to a probability distribution.
Watermarking just alters the pseudorandom number generator. If "I like turtles" was previously the response to your prompt with probability 100%, it will still be so. This is why watermarking is only effective for long strings of text
It's like the sudden change of a language style and its verbosity didn't happen recently.
To random words you pick and provide a sufficient amount of text to vary with random number without losing its meaning you need a text with high entropy.
Nothing about watermarking would require padding the response length with pseudo-intelligible Claudese. Regular filler would work fine.
Also, it would probably provide higher entropy to write normal human-sounding English instead of reusing a repetitive grab bag of load-bearing phrases. This theory doesn't really make any sense.
No, it basically uses a fixed seed for the random number generator to generate tokens. The generated random sequence is just as random as it would be with a random seed.
The ultimate goal might be some political aim and in that sense cruelty is not the ultimate point but what is usually meant by this phrase is that the cruelty is not an unintentional oversight, an unintentional side effect. It is intentional (and there to achieve the ultimate political goal).
I totally get your thought process, especially with the dry exercise description in the comment you responded to. That’s maybe an approach for some people but it also doesn’t seem that sustainable.
My view on exercise is that it should give me something in the moment. That also helps keep me motivated.
So arduous hikes and bike tours aren’t something I do for the exercise, that’s just a nice side benefit. I do those because I want to get to exciting places or explore the world at a pace that makes sense.
I ride my exercise bike to keep in form and be able to do those hikes and bikes tours, even to go further and enable me to reach new places and do new things. Same with swimming, though that also helps me to clear my head and think about stuff. Both of those can potentially be boring, but at least I always feel amazing and energized afterwards.
Strength training is admittedly something I do almost exclusively for its long-term health benefits (and struggle the most with), though I also like feeling stronger and seeing my body change at least a tiny bit. Oh, and it’s an activity (as well as some Yoga and flexibility exercises) I do together with my partner, so it always has that social component.
Bouldering and badminton (both of which I absolutely suck at) also have that social component: it’s something I do with other people.
Oh, and then I don’t own a car, so my bike is my only means of getting from a to b and transporting things. So I automatically get moving that way and it never feels like wasted time or even exercise to me.
That’s much more than five hours per week spent exercising and very little of it is boring or feels like empty time that is merely an investment into my future. I do think that’s the trick to being and staying active. At least for me.
Good and best is doing a lot of heavy lifting in your comment.
What are you talking about when you speak of good/best outcomes? I think especially when it comes to talking about fitness it is important to be explicit about outcomes because only with outcomes in mind can you talk about what “good” actually means.
In that context I’m a bit confused by your focus on (endurance) athletes. Their overriding goal is obviously increasing their athletic performance in a way that keeps them injury free. Their main focus aren’t health outcomes. That’s not the same. So if you focus on them you are in danger of talking past each other.
To that point: sure, those studies often have unrealistic protocols but the way (performance) athletes train is also unrealistic for most people!
If those studies are about population health they should have realistic protocols (in terms of: realistically achievable and adhered to by non-athletes, in the best case also empirically grounded in the sense of: this is a realistic range of activities non-athletes might actually engage in on a regular basis) for non-athletes.
Political change is needed to achieve any of this. Changing your individual behavior or even the individual behavior of everyone you know – while certainly not wrong! – won’t solve this.
This is like trying to abolish slavery in 19th century US by not owning slaves and trying your very best to not buy anything that involves slave labor … laudable, but not a way to abolish slavery.
I think that’s an interesting point and one I did actually think about while writing my comment but ultimately didn’t address.
The interaction between individual behavior changes and political changes is interesting and to my mind where lots of interesting stuff happens. I’m not 100% convinced that politics is always downstream from culture and culture is downstream from individual behavior changes but I do agree that there is an interesting interaction happening that’s worth looking at and addressing.
The abolitionist in isolation certainly will not be able to abolish slavery, however their individual behavior might contribute to an atmosphere where slavery is ostracized and seen as more obviously evil by more people, thus making it easier for politicians to make political decisions that ultimately lead to the abolition of slavery.
I think that is why it is always so important for me to point out that it is certainly never wrong to individually behave in a way the limits your own individual contribution to climate change. I mostly have a problem when that‘s the end or even the main focus of it. Climate protection to my mind is most effective when it is tightly coupled to clear political demands. Too often individual behavior changes are used as action terminators. My impression is that individual action is often presented (as here) as a complete solution to climate change when it is anything but.
Exactly! Politics try to appease the mass. If the mass would be willing to put climate change first we would have other policies, instead we follow the dark visions of lunatic billionaires
All I know is that they worked hard on this game and found plenty of people who enjoy playing it a lot and that’s that. Doesn’t mean they are wrong, doesn’t mean I’m wrong in bouncing off. I do think it is interesting to explore why I bounced off and others might not – I can identify with many of the negative comments here, especially the 10,000 bowls of oatmeal one regarding the procedural generation which probably was the one thing I was most fascinated by when I decided to install the game – but there are so many other systems and elements in this game that if you more into those I can easily see why the game works much better for you.
reply