Hacker News Reader

Hacker News

Comment by vonneumannstan | Hacker News Reader

Parent Full thread

vonneumannstan

[flagged]

View on HN

Depends on your perspective.

If you are the owner of one or more AI companies, it could be very good for you.

How about more of your thought process and reasoning instead of a blanket ad hominem attack?

That's not an ad hominem attack.

Calling someone a name is not an ad hominem.

Using something about someone to undermine an argument is an ad hominem.

People against AI regulations are lunatics: Not ad hominem. That guy is a lunatic so his opinion about AI regulations are also insane: Ad hominem.

You’re saying the same thing twice. One just more blatantly than the other. The point of calling someone a lunatic is to undermine their opinion. It’s ad hominem either way even if one is more subtle.

> The point of calling someone a lunatic is to undermine their opinion.

I don't see that as the point. I think they were just opining. There doesn't seem to be an argument they are making or refuting so this cannot, by definition, be an "argument ad hominem" (argument to the person).

The parent comment to yours nails it with the given examples. The first is the conclusion to arguments -- which they did not expound on -- "this opinion is lunacy"; and the second is specific argument that "you are a lunatic" and conclusion that therefore "your opinion is lunacy". That is an argument to the person.

(Not to speak to the quality of OP's comment for this message board. A comment explaining why they think it's crazy would be preferred.)

"anyone who arrives at the conclusion you have arrived at is a lunatic" is functionally the same as "you're a lunatic so your opinion is invalid". Nitpick all you want.

"you are a lunatic BECAUSE you're opinion is invalid" (translates to: you are being unreasonable). Not ad hominen.

is different from

"you are a lunatic THEREFORE your opinion is invalid" (translates to: I think your conclusions came from your mental condition). Ad hominen.

It's quite easy to understand.

Fallacy lists are known to be vulnerable to sophists that take it apart literally instead of seeing the broader perspective of having them.

The introduction of a disruptive discussion (such as: is this a fallacy or not?) is a common technique for diversion.

You should have acknowledged that the other guy didn't meant to insult when he tried to explain. But you did the opposite, you doubled down on making it an insult. It's a sign of a bad conversation, in which one or more participants are not interested in the discussion.

I disagree. Using such a broad definition means any negativity during a discussion can be called out as an "ad hominem" (and it happens all the time).

No. It's saying that personal attacks that are meant to undermine the credibility of your opponent is ad hominem. Because it is.

Do you really think having 50 completely different regulatory regimes is somehow better?

This isn't so bad. Also, since when did HN become pro-tech-regulation (especially pro-emerging-tech regulation)? It is a weird change.

HN shifts when tech becomes dangerous.

There has been a shift against destructive social media and addictive phones too.

Idk what Libertarian dreamland you people live in. Go try and live in a country with no regulations for things like pollution, hazardous waste, aviation safety, etc, etc and let us know how long you last. Powerful technologies necessarily can be dangerous. AI is going to be a major change driver and honestly if you deny that it's an incredibly powerful technology then I don't know what else to say.

On the contrary, I often mention here just how transformative I expect it to be. I think AI is still being underestimated by a lot of people here.

That's exactly why I want more people to have it and more smaller companies to be able to build things with it.

50 different regulatory regimes will make that much harder and lead to more centralization. This isn't even libertarian. I don't care if the government has good regulations, but diffuse local regulations for national companies is not the answer.

> Do you really think having 50 completely different regulatory regimes is somehow better?

Yes, that is a less bad problem than a law that bans any enforcement of such regulations. 50 different sets of regulation just means that companies will change policy to adhere to the most strict of them (presumably California's) and be done with it.

Completely agree, 10 years is an eternity in AI development. AGI will 100% be here before then.

that's a very confident %

If you read the article, it is clear that it doesn't block any and all regulation of AI. It says states cannot make federal funding follow non-federal rules around AI. The federal government may actually have more regulations than states, and this would require states to do a better job.

Now you get a grey area when AI is being added to everything to use as an excuse to avoid state laws

Read harder. There is zero connection to receipt of federal funding.

https://d1dth6e84htgma.cloudfront.net/Subtitle_C_Communicati...

IN GENERAL.-Except as provided in paragraph (2), no State or political subdivision thereof may enforce any law or regulation regulating artificial intelligence models, artificial intelligence sys-tems, or automated decision systems during the 10-year period beginning on the date of the enactment of this Act.

It’s an across the board blanket preemption.

There's no indication the Feds are going to take any initiative on this at all. They lap up Altman's pandering at Senate Hearings and will do nothing.

> The federal government may actually have more regulations than states

There would be a chance of this being true if the language didn't bar states from enforcing their own regulation; there's no point to that except for a worry that some states' regulation will be more strict.

It's also very clear this doesn't pass Tenth Amendment muster.

(Or shouldn't, at least.)

AI has significant impacts on interstate commerce, which the feds have basically carte blanche to regulate.

Reply on news.ycombinator.com