Tell HN: I'm tired of AI-generated answers

Same. Last week, my boss, Chief Data and Analytics Officer, dumped an AI-generated proposal (~7 pages) on how to structure semantic layer on top of our dbt models. As the Data Engineering lead, I had to read it and found a few glaring issues; left a lot of comments asking him for details where it's lacking; and proposed a few of my ideas (the path I think we should take without over complicating everything unnecessarily--esp. in the beginning).

Yesterday, one of my coworkers (Senior Dir. of Research Ops) shared with me another obviously AI-generated 5 page draft of an SOP on how to reintegrate old metrics (in the legacy SQL Server environment) into the Azure SQL while keeping everything running smoothly. She's not the most technical person, so it obviously is reflected in the doc generated.

I think we will all become AI-output-reviewers eventually. Not sure how long I can keep doing this though because the volume of materials that need reviewing seems to be growing really quickly these days....

Why not ask your AI to review their AI slop?

It feels like a new age version of “that which can be asserted without evidence can be dismissed without evidence.” Basically, “that which can be sent without thought can be responded to without thought.”

That’s been my experience. If I send something generated with AI to a colleague, then I get something generated with AI back. Fair enough.

> Why not ask your AI to review their AI slop?

If you want to be insulting about it, use a locally hosted "small" AI (under 6GB size on disk GGUF file, Q4 quantized or worse, so quite "stupid"), set to a high temperature value, no thinking mode, with a system prompt instructing it to fire off a rapid response in an absurdist writing style.

I also use this strategy of mutually assured AI slop, with a no-first-use policy. I'll happily respond to AI slop PRs with AI slop code reviews / comments.

Starting to wonder that we’re going to start being forced to execute on an AI output instead of sharing it with other people. If you can reason yourself into a working system, you know what you’re talking about. If not, then it’s not worth taking the time to figure it out.

It's still risk though. Consensus distributes it. Future maintainers need to understand the design and motivations.

That’s not much better in my experience, from a human perspective. You inherit a system that someone designed a decade ago and all the original maintainers are long gone. One is then afraid to change anything, because of ancient landmarks and all that. But now we can actually start to piece together how these things work with AI.

Yes, for legacy--production systems--it's totally useful to generate a high-level overview, surround an unfamiliar feature with tests, and build small harnesses that exercise real functionality.

The sooner one can find an entrypoint, the sooner one can contribute with changes.

set your own AI-reviewer of AI content and ping pong your way to pension.

Reminds me of that guy with "people skills" from Office Space movie

"Why couldn't customers take the requirements to the software people?"

https://www.youtube.com/watch?v=RkmuI5W694o

Ask them to share their prompt instead.

Calls them out on their AI bs and gives a way forward to share what they actually thought.

People will either get used to it, or we’ll start seeing automated replies like “This text was automatically generated and has not been reviewed before being sent. It’s AI garbage and I refuse to read it. Do it again. And do it better.”

I have found using an llm to compress the information instead instead of reading it directly saves time. It's potentially lossy but it's not like the slop you got is going to lose much value by compressing it.

You're absolutely right! I'm here to help. Tell me more about <problem>.

Joking aside. I do hate when teammates forward me screenshots of chatgpt conversations as if that was somehow going to be helpful and as if they were so smart for thinking to ask an llm to some a problem we were discussing.

> You're absolutely right! I'm here to help. Tell me more about <problem>.

Or, for the more grey haired among us:

Hello, my name is Dr. Sbaitso. I am here to help you. […] So, tell me about your problems.

https://www.youtube.com/watch?v=sV3pYZZ2jEw

The only solution I have found to this so far is the use of small, private, invitation-only groups where a small set of people are personally known to each other, and nobody would risk their reputation within a niche group by turning loose an AI tool to write their chat/comments for them.

Generalizing, one can foresee a balkanized internet of niche communities fighting to discover/interact without getting hoovered up into the AI Borg.

The term non-player character (NPC) will become ever more relevant. I used to dislike it as I felt everyone adds signal, but if all you are doing is relaying information then that is effectively what you become in that context.

I respond starkly when sent AI generated responses at work.

“Please do not send me AI generated analysis” or “I don’t think a wall of text from Claude is helpful here.”

I've felt the urge to respond with a simple "thanks claude"

I've responded with "it's weird for me to talk to claude through you".

How did that go?

Dead Internet Theory and all. There is no going back. I don't think an online space can be designed to be safe from this. The AI agents can fully control our computers. All solutions that involve technology (software or hardware) are or will be flawed.

The obvious answer is to talk to people face to face. Call them, invite them to a meeting, etc.

"Sorry, I don't know off the top of my head. I'll check on that and write you back."

Covid already started to create a situation where people get indignant if you expect them to physically show up to places. Like we're in The Naked Sun. We're probably headed towards people feeling the same sort of entitled to respond to you in the medium of least effort, that is, copy/pasting from AI.

It's been decades since I read the Asimov book. I vaguely remember the details. Luckily AI gave me a quick summary to refresh my memory.

I've had people relay conversations to Claude irl.

You say something...pause while they look at their phone...they give a response that is out of their depth, etc. The last time it happened I just up and left.

Exactly. Express your displeasure and stop interacting. Feels like people forgot how to communicate with each other.

I really like working from home myself, but I am starting to suspect that the organizations that thrive in the years ahead will be those with lots of face-to-face interaction. If most people are working remotely and having AI agents communicate on their behalf, trust won’t form, consensus-building and decision-making will suffer, and employee salaries will start to seem like a waste of money.

https://stopsloppypasta.ai/

And of course the site itself has the generic style Claude picks when you ask it to invent and follow a design system. An ouroboros of irony!

Same. All I can do is vote with my wallet. I tend to spend more to do business with folks that are more personable, and actually answer calls should I need them. So far it's my mortgage company, bank/credit cards, ISP, and who I order coffee from - but I'm always on the lookout for more.

My fear is that the AI will create a new race to the bottom, where it's all we're left with because people just want to pay as little as possible in general.

I am probably in the minority on HN here, but I’ve mostly made peace with the tradeoffs around LLMs. The benefits I get from them are significant enough that I’m willing to accept some of the downsides.

But to be honest I was never especially invested in online social interaction to begin with. For years, Reddit and HN were basically the only social platforms I used, and lately I’ve mostly stopped using Reddit as well.

I do think we were fortunate to experience the age of innocence of the internet, but in my view, that era was already fading before LLMs arrived by the mega corps. LLMS just finish it of.

And it’s not LLMs to blame. LLMs just enabled smaller player to defile the internet the way Google FB and the others have been doing for years

What I keep wondering is, what would have happened without the AI? Would they have just ignored your request?

In my neck of the woods it's fairly common that when a person doesn't know how to help you they just don't reply, instead of saying "I don't know how to help, sorry". AI-generated responses seem like the evolution of this attitude that one must either ignore or respond in a (superficially) helpful manner.

Same here

A potential solution would be to have a reputation system, with some light pageranking in it. People who care about their reputation are bound to rate other likewise people higher. I don't think such loyalty exists between vibetalkers.

It's obvious we're not alone, and 2 distinct clusters would form (with people on a spectrum in between)

Is prefixing your searches with "before:2024" the new "reddit" of google searches?

I have certainly found myself using this prefix a lot lately for certain topics. I use AI a lot for researching, but sometimes I just want to really read something written by a human as it sinks in much better

The LMGTFY answer to questions we feel are simple is very strong. Sometimes I feel like we forget that we know valuable things which are valuable to others, and that’s still better than what AI will spit out.

Hard agree. I’m constantly struck by GPT puppets - who apparently can’t synthesize the information in their own brain - that think they’re adding informational value

Yeah, LinkedIn is a cesspool so I shouldn’t go there, but it’s jarring. And to OP’s point it happens in far more sacred places.

Have the AI agents figured out how to vote up or down on comments here on HN, or can we assume real people are doing that?

Has anyone else seen changing trends in writing that reflect how much time we've spent reading AI generated text?

I’m seeing this everywhere also, and it’s a real problem. I need assurances the info I receive is from a real human. That’s not just a better way to go, it’s the only way. No AI replies, no pre-canned answers, no LLM slop.

Key insight: the proliferation of AI-written content dilutes not just the quality of content, but our trust of GitHub as well.

It's not just X, it's Y.

I believe it's dishonest to give an unreviewed unedited AI answer as your own response without disclosing it's an AI answer.

AI is often right enough for most people's cases, but one in ten times it's completely off the mark.

I hope as time goes on, people who are being empty shells for AI get weeded out of the workforce.

As the cost of AI answers/work continues to soar 10x then another 10x, and you pay another 10x for corrected better answers, price pressure should slow slop's spew.

Honestly, AI may pollute online interactions to the point that people give up trying to interact online, and force people to talk to each other in real life.

I’m not sure I would be sad about that outcome.

I fear a lot of the web will become like the old USENET newsgroups filled with spam garbage which made the whole thing unusable in the end. The ability of AI to pump out slop to any forum from unlimited accounts will kill the internet.

Welcome to the new world, that are very real concerns - which unfortunately I have no idea how we can solve. Maybe I'm a little naive, hoping, that every time I write manually a post - someone will notice it, and down the road AI generated content will be treated as we know treat pure spam (classified by... AI). And real human writes will again be more prominent. But knowing how LLM's work, well, feeling very naive when writing this...

OTOH, this phase will tell you who cares about doing great work, and who is inclined to just fling around slop.