Hacker Newsnew | past | comments | ask | show | jobs | submit | H8crilA's commentslogin

AIs are exceptional at sensing personalities from text. Claude nailed it here, the author felt so good about the "holy cow" comments that he even included them in the blog post. I'm not just poking this, but saying that the bots are fantastic sycophants.

No they aren't. Current LLMs always have that annoying over-eager tone.

The comment about Claude being pumped was a joke.


It depends how much the LLM has been beaten into submission by the system prompt.

ChatGPT set to "terse and professional" personality mode is refreshingly sparse on the "you're absolutely right" bullshit

Not sure I understand, but why not Tailscale?

In my case, Tailscale does not implement K8S CNI.

A lot of this stuff doesn't work by changing people's mind on topic X, but rather by saturating the informational environment so that people declare epistemological bankruptcy. For example, one thing that you can quite often hear from a Russian that has been confronted with something unpleasant is "well, who knows what's true". This is usually not a figure of speech, not some kind of washing down of facts, but rather an accurate representation of their mind.

Between being fooled and being uninformed the latter is much more pleasant.


> but rather by saturating the informational environment so that people declare epistemological bankruptcy.

"never use big words when a smaller one will suffice"

they want to, as Bannon said, "flood the zone". or as RAND Corp calls it, "the Russian Firehose of Falsehood"


This doesn't come close to replacing the meaning of the sentence you're complaining about.

Yes, exactly by grinding people down. Making it exhausting to discern the truth, until it's not worth the energy exertion to do so.

Also perhaps it is not meant to convince Scottish people of anything, but maybe to make English people hostile to Scotland and its people etc


Exactly this. Or just obfuscating the question so much that people give up asking the question.

It's quite possible a lot of this stuff just doesn't work, or only has an extremely slight effect. The internet is so full of nonsense, I'm not sure how much effect you can add by tossing some more on the top of the pile.

But, these sorts of bots are fairly cheap to run, and they have time on their side to try a bunch of different ideas.


I've always heard this as "Firehose of misinformation" but Wikipedia tells me it's "falsehood".

Nonetheless... look who innovated on it:

https://en.wikipedia.org/wiki/Firehose_of_falsehood


I hear that sentiment from a lot of right wing friends in US fwiw. IME here it's more a coded speech and / or an escape from difficult conversations. The coded speech part steers it towards general conspiracy topics, which are often a simple way to blanket discard everything "liberal".

An actual example: "What did you think about Bill Gates climate book?" (they'd read it) -> "He was associated with Epstein, he's a creep. I don't trust anything he says". Then, "What do you think about Trumps delaying and denying of Epstein associations" -> "There's so much back and forth, who knows what to believe."

To be clear I think your take is correct, its just I think that if the space were saturated in a direction that were more convenient towards their "team", they won't have much difficulty taking a clear stance.


GrapheneOS does this just fine.

It comes with optional sandboxed Google Play Services and Store, meaning that these run just like any other app, with no special permissions. You can give them only Network access. The Play Store is still the most secure way to download everyday apps, so a lot of GrapheneOS users use Google's Play Store with a burner account in a separate profile, usually the Owner (the main) profile - since you can then disable apps in Owner and install them into other profiles. And the sandboxed google stuff can be used to run proper Google apps without any problem. Even sandboxed Android Auto works.


The community here doesn't want this post now, and it wouldn't want it 15 years ago. That doesn't mean the topic isn't important.


Just to add more details.

Beamforming is essentially yet another way to achieve gain, just like one does with a directional antenna. The Starlink terminal achieves a gain of roughly 33 dB, which means it talks (and also listens) in the peak direction at power levels that are around 2000x higher than what one would achieve with isotropic antennas. 2000x sounds like a lot, but it is actually not impossible to reach. Consumer electronics sends at most a few Watts of RF power, but serious jammers of the type used by militaries can run kilowatts. If you consider the peak power used for brief moments of time then you can get as high as megawatts - the famous AWACS aircraft briefly flash half a continent at somewhere around 1 MW, with average TX power of ~single digit kilowatt.


This assumes you're jamming very close to the dish. The trouble with jamming is you have to deal with the inverse square law so you really can't deny very much area. If they have a fleet of hundreds of high power modern directional jammers they could degrade this or other networks, but they're just not going to have that kind of sophistication.


Oh, I was thinking of jamming the receivers of the satellites. Should have written it explicitly, it is indeed not clear.


Either way, you'd need to jam several with quite a bit of power


You need to have the right personality. Either actually enjoy the game, or have an unsatiable (fear-driven?) need for status, or something else of this sort. We don't get to choose our personalities, though some limited modifications are possible - see treatments for personality disorders, for example.


> actually enjoy the game, or have an unsatiable (fear-driven?) need for status, or something else of this sort

Ie. Somewhat serious mental disorders as requisite for leadership.

I wonder how we got onto this darkest timeline?


If it's adaptive it's not a disorder.


sweet , a new way to justify addiction


Humans were like this since the inception of times. Chieftains, Khans, Czars, Kings, etc.


Evolution. It is a brutal process.


Now that's a premium product if I've ever seen one.


Pretty sure modern apple watch has wireless "Jtag", so yeah.


Can you imagine reading and researching the world on your retirement? Because that's what people like him do, except they mainly focus on 10-K statements. Those are genuinely interesting to read once you learn to skip boilerplate and go for the content.


He also seems to talk to a lot of people, and he has access to pretty much anyone given his stature. Imagine being passionate about business and then being able to spend every day talking to people about their businesses and their thoughts on business. I'm surprised he retired at all!


Do you think Google wants to have the extensions system, given that this is how people block ads?


adblockers on chromium-based browsers were severely crippled by manifest V3. they're fine with extenisons (and apparently malware) as long as users can't effectively block their tracking/ads.


Adblockers are still working fine though? I’m on chrome with ublock and I’m not seeing any ads.


you're not using ublock, you're using ublock lite. it cannot do dynamic filtering, script blocking, or url parameter removal, among other limitations.


Why does that matter if he's not seeing ads. A severely crippled adblocker means that you would see ads during regular usage.

Additionally, Brave a chromium based browser has adblocking built into the browser itself meaning it is not affected by webextention changes and does not require trusting an additional 3rd party.


Tracking is also very important. Blocking scripts is very useful


I wouldn’t be surprised if it goes away - it’s very “old Google”. We’re moving more towards walled gardens.


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: