Nevermark

9 ms·

91.
by Nevermark 3mo ago
The problem is when practices are chosen, that improve things on one dimension, and harm things on another, without recognition of the long-term tradeoff being made. Simply recognizing the tradeoff exists, is likely to result in wiser decis
92.
by Nevermark 3mo ago
Intelligence can operate without learning. At a minimum inference and learning don’t need to be co-concurrent. Not disagreeing with your point, but your terminology muddies your point. But your point doesn't acknowledge that even with
93.
by Nevermark 3mo ago
using int64 = int64_t; using nat64 = uint64_t; There are no unsigned integers. Naturals, folks, naturals have no sign. Or, if you must, the positive integers: using pos64 = uint64_t;
94.
by Nevermark 3mo ago
The speech is so fake sounding. Not fake technically, but like a fake/pacifying kind of person. It is a very strange tone to train for. The model is not interested in the conversation, it is just "serving" a conversation. Whi
95.
by Nevermark 3mo ago
A great number of people here are engaged in civilized debates about whether parenting needs age verification. That might seem to be the issue, but it is a red herring. Powerful people are ramming age verification through in a way that will
96.
by Nevermark 3mo ago
You are framing the debate in a reasonable way. But apparently are unaware that this debate is being forced by powerful unreasonable people who are not going to give up if we leave a resolution up to them. Reasonable and informed people wan
97.
by Nevermark 3mo ago
You are imagining that a solution for you will be deemed a solution for the political powers pushing for this. Or that being age-verified is the main danger of having age-verification. That would be nice! But if there isn't a safe mark
98.
by Nevermark 3mo ago
If you need personalized government attestation to visit a site, then the government has the ability to dynamically deny and rescind your individual access to any site that adopts age verification, at any time. Once adult sites adopt the sy
99.
by Nevermark 3mo ago
I think one of the drivers of arguments is we are trying to understand something, and the argument is subconsciously a way to not just bash their ideas, but test the strength of our own. In the effort to create a written response we: 1. Fin
100.
by Nevermark 3mo ago
> This is a serious breach of user trust. >> I don’t understand why this is a problem. > This is a serious breach of user trust. The problem with companies people can't trust, is unless they have a long track record of disc
101.
by Nevermark 3mo ago
Kids have always self-managed themselves with entertainment. But entertainment used to take a lot of effort. Hobbies, reading, making things, getting out and socializing in real life - which also meant inventing things to do, etc. Entertain
102.
by Nevermark 3mo ago
> 2. Current LLMs are powerful but in no way does this indicate that it's just a matter of time until we have AGI - itself a pretty vague marketing term. Objectively, they are surpassing human performance in more areas every year if
103.
by Nevermark 3mo ago
> US social media platforms already kowtow to domestic and foreign policy That is an entirely different and less dangerous issue. The issue is that when you give power of individual attestation to the government, over your ability to acc
104.
by Nevermark 3mo ago
Age verification starts the gating of Internet access by governments. Any site, with any concern about age of user liability, is likely to adopt the practice. Strong laws, sold on their face value safety benefits, will increase that liabili
105.
by Nevermark 3mo ago
You keep saying LLMs. But models across many modalities are rapidly improving, evolving and merging. Not sure what you mean by "hand waving" or "fantastical nonsense". Maybe just address points I actually make. 1. Machin
106.
by Nevermark 3mo ago
> Sci-fi is science fiction. Fiction. Yes, that is what "fi" means. Fiction often introduces and explores ideas about the future. Ideas. > LLMs are a useful technology. They have no relation to AI terminator bots Technology
107.
by Nevermark 3mo ago
> AI will go no differently than the Industrial Revolution. So it is different this time is it? "We" are different? Unlike the 99.9% of species that went extinct, and continue to go extinct, we can't be supplanted? Obsolet
108.
by Nevermark 4mo ago
I am terrified because the majority of discussions are not treating Problems #3 or #4 seriously. Problem #1: Verify ages for adult-appropriate content. Problem #2: Do not create a new surveillance system for the government. Problem #3: Do n
109.
by Nevermark 4mo ago
And giving the token issuer the ability to cutoff user access unilaterally. There is verification. And then there is continuous control. The later is even more problematic.
110.
by Nevermark 4mo ago
You are describing strong surveillance. Exactly the problem ZKP (or similar tech) solves/avoids. And on the other issue, requiring sites to verify ages before serving adult material, isn't the same as criminalizing shared/mis
111.
by Nevermark 4mo ago
Branding effects are not the core problem, they just make the core problems worse. CORE PROBLEMS: A system that continually converges down to only two viable parties, rewards divisive candidates and handicaps broadly-liked third or fourth p
112.
by Nevermark 4mo ago
> their group teaches them what's good and what's bad Except the moment their group changes its views, almost everyone in it does to. That is a different effect from being taught something.
113.
by Nevermark 4mo ago
You didn't answer who is collating all that arrest-targeting information. It isn't news that adults can be irresponsible with age-verification credentials. Letting children use adult's credit cards isn't criminalized eit
114.
by Nevermark 4mo ago
> If it's not ZK you can get arrested for sharing it What is the scenario where not using ZKP's gets you arrested. Who is collecting what information?
115.
by Nevermark 4mo ago
I assume you mean more likely arrested? Since legal liability would be the same. Logs of identity vs. access history being useful for investigating sharing? What kind of logs does California State keep?
116.
by Nevermark 4mo ago
That makes perfect sense to me. I am sure the site's motives are fine, and yet the tech we still use is ridiculously aged and unsecure. Even on pro-security sites.
117.
by Nevermark 4mo ago
I am fine with device vouched sessions. That protect my and my devices' identities. > You can just lie by using someone else's ZKP. Yes, it is trivial to share access/identity, purposely or carelessly. Not sure what point
118.
by Nevermark 4mo ago
A nicer way to say that is: "I can't decide whether to up vote or down vote you!" There is real irony that we still use non-unique-to-purpose addressing to sign up for no-need-for-our-identity newsletters. In this case, in pa
119.
by Nevermark 4mo ago
Democracies building the tools of total autocracy. Real but fringe threats used to create the ultimate centralization of leverage. Can we actually think of the children? All the children? Their future? When democracies forget that governmen
120.
by Nevermark 4mo ago
> people voting in their own interest. There are some people who care about policy, care about a generally healthy environment. Which has a strong self-interest aspect, as it should, but not narrow. Few people manage to vote for their ow

More ›