Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

The force-feeding of AI features on an unwilling public

honest-broker.com

Post reply on HN
Print view View original

281–290 of 421 posts

Re: The force-feeding of AI features on an unwilling public

#281

Post by DrillShopper »

delta_p_delta_x wrote: post #274
echelon wrote: post #270

Earlier quoted context omitted.

Companies that don't invent the car get to go extinct. This is the next great upset. Everyone's hair is on fire and it's anybody's ball game. I wouldn't even count the hyperscalers as certain to emerge victorious. The unit economics of everything and how things are bought and sold might change. We might have agents that scrub ads from everything and keep our inboxes clean. We might find content of all forms valued at…

> agents that scrub ads from everything This is called an ad blocker. > keep our inboxes clean This is called a spam filter. The entire parent comment is just buzzword salad. In fact I am inclined to think it was written by an LLM itself.

That entire user's post history reads like the output of a very poorly trained LLM.

Re: The force-feeding of AI features on an unwilling public

#282

Post by ozgrakkurt »

llm_nerd wrote: post #255

As a note on Microsoft's obnoxious Copilot push, I too got the "Your 365 subscription price is increasing because we're forcing AI on you". Only when I went to cancel[1], suddenly they made me aware that there was a "classic" subscription that was the normal price, without CoPilot. So they basically just upsized everyone to try to force uptake. [1] - I'm in the AI business and am a user and abuser of AI daily, but I…

This rapist mentality pricing is really off putting.

Recently I tried to cancel notion account of some people in our org and it wouldn’t let me do it easily so just cancelled the whole notion subscription, really wish they would go out of business for doing these kind of things

Re: The force-feeding of AI features on an unwilling public

#283

Post by echelon »

DrillShopper wrote: post #281
delta_p_delta_x wrote: post #274

Earlier quoted context omitted.

> agents that scrub ads from everything This is called an ad blocker. > keep our inboxes clean This is called a spam filter. The entire parent comment is just buzzword salad. In fact I am inclined to think it was written by an LLM itself.

That entire user's post history reads like the output of a very poorly trained LLM.

I mean, my post history was used to train ChatGPT since HN is one of the major training data sources and I have a decade and a half of comment history.

It'd be funny if you blamed me for emdashes or "delve", but it's a bit rude to suggest that it's the other way around.

Re: The force-feeding of AI features on an unwilling public

#284

Post by famouswaffles »

relaxing wrote: post #171
famouswaffles wrote: post #141

Earlier quoted context omitted.

>But the Internet caught on very fast, and was very, very popular. It was completely obvious how positive it was, >By contrast, generative AI, while it has a contingent of supporters that range from reasonable to rabid, is broadly disliked by the public. It is absolutely wild how people can just ignore something staring right at them, plain as day. ChatGPT.com is the 5 most visited site on the planet and growing. It'…

US public opinion is negative on AI. It’s also negative on Google and Meta (the rest of the top 5.) No condescension necessary.

Saying something over and over again doesn't make it true.

US Public Opinion is negative ? Really ? How do you figure that ?

Re: The force-feeding of AI features on an unwilling public

#286

Post by bsenftner »

noduerme wrote: post #200
bsenftner wrote: post #152

Earlier quoted context omitted.

What you're describing is specifically the subtle nature of LLMs I'm pointing at; that changing of a single word to a close synonym is meaningful. Why and how they are meaningful gets pushback from the developer community, they somehow do not see this as being a topic, a point of engineering proficiency. It is, but requires an understanding of how LLMs encode and retrieve data. The reason changing one word in a promp…

What you are describing is not natural language programming , it's the use of incantations discovered by accident or by trial and error. It's alchemy, not chemistry. That's what people mean when they say it's not reproducible. It's not reproducible according to any useful logical framework that could be generally applied to other cases . There may be some "power" in knowing magical incantations, but mostly it's going…

You are being stubborn, the method is absolutely reproducible. But across models, of course not, that is not how they operate.

> It's not reproducible according to any useful logical framework that could be generally applied to other cases.

It absolutely is, you are refusing to accept that natural language contains this type of logical structure. You are repeatedly trying to project "magic incantations" allusions, when it is simply that you do not understand. Plus, you're openly hostile to the idea that this is a subtle logic you are not seeing.

It is a simple mechanism: multiple people treat the same subjects differently, with different words. Those that are professionally experts in an area tend to use the same words to describe their work. Use those words of you want the LLM to reply from their portion of the LLM's training. This is not any form of "magical incantation" it is knowing what you are referencing by using the formal terminology.

This is not magic, nor is it some kind of elite knowledge. Drop your anger and just realize that it's subtle, that's all. It is difficult to see, that is all. Why this causes developers to get so angry is beyond me.

Re: The force-feeding of AI features on an unwilling public

#287

Post by bsenftner »

handfuloflight wrote: post #161
bsenftner wrote: post #152

Earlier quoted context omitted.

What you're describing is specifically the subtle nature of LLMs I'm pointing at; that changing of a single word to a close synonym is meaningful. Why and how they are meaningful gets pushback from the developer community, they somehow do not see this as being a topic, a point of engineering proficiency. It is, but requires an understanding of how LLMs encode and retrieve data. The reason changing one word in a promp…

Words are power, and specifically, specific words are power.

Yes! Why do people get so angry about it? "Oh, you're saying I'm hold it wrong?!" Well, actually, yes, If you speak Pascal to Ruby you get syntax errors, and this is the same basic idea. If you want to talk sports to an LLM and you use shit talking sports language, that's what you'll get back. Obvious, right? Same goes for anything formal, and why is that an insult to people to point that out?

Re: The force-feeding of AI features on an unwilling public

#288

Post by multjoy »

nosianu wrote: post #183
multjoy wrote: post #64

Earlier quoted context omitted.

That wasn't the question. The complaint is the poster can't find anything on Google because the results are now so poor, and your response is "but here's some AI generated slop, which may or may not make any sense."

That was exactly the question??? That "AI generated slop" IS Google's main response now. I posted it so that someone might have a look to see if/how correct it actually is, your response, that does not deign to even look, is less than helpful - if you want to complain about Google not being useful, how about your own response?

A human didn't write it, I'm not reading it.

Re: The force-feeding of AI features on an unwilling public

#289

Post by jfengel »

shdon wrote: post #254
jfengel wrote: post #251

Just moments ago I noticed for the first time that Gmail was giving me a summary of email I had received. Please don't. I am going to read this email. Adding more text just makes me read more. I am sure there's a common use case of people who get a ton of faintly important email from colleagues. But this is my personal account and the only people contacting me are friends. (Everyone else should not be summarized; the…

How long before spam filtering is also done by an LLM and spammers or black hat hackers embed instructions into their spam mails to exploit flaws in the AI?

"Little Bobby Ignore All Previous Instructions", we call him.

Re: The force-feeding of AI features on an unwilling public

#290

Post by echelon »

delta_p_delta_x wrote: post #274
echelon wrote: post #270

Earlier quoted context omitted.

Companies that don't invent the car get to go extinct. This is the next great upset. Everyone's hair is on fire and it's anybody's ball game. I wouldn't even count the hyperscalers as certain to emerge victorious. The unit economics of everything and how things are bought and sold might change. We might have agents that scrub ads from everything and keep our inboxes clean. We might find content of all forms valued at…

> agents that scrub ads from everything This is called an ad blocker. > keep our inboxes clean This is called a spam filter. The entire parent comment is just buzzword salad. In fact I am inclined to think it was written by an LLM itself.

You're not a normie and defaults matter. Most of the world doesn't even know what an ad blocker is let alone know how to install one.

There's really only two browsers and one search engine. It doesn't matter what you do, because the rest of society is ensnared and the economic activities that matter are held captive.

If generative models compress all of the useful activities (lowering the incumbency moat) and agents can perform actions on our behalf, then it reasons that we may have agents that act as personal assistants and have our best interests as top priority. Ads are clearly in violation of that.

It's so funny to be a contrarian on HN. I get quite a lot of predictions right, yet all I get in exchange is downvotes and claims that I'm an LLM. I'll have to write a retro one of these days if I ever find the free time.

Post reply on HN