I refuse to date someone who uses ChatGPT
theguardian.com
theguardian.com
Do they have a similar reaction to searching "restaurants near here" in a maps app? Practically any time I go out to eat I'm either going to one or two favorite places, or I'm looking on the internet where something gives me a list of possibilities.
It's not like I've memorized every local restaurant, so looking online for recommendations online is a way to learn about new places.
To suddenly go from "this is perfect" to "ick" is a problem with being too judgmental and not self-regulating.
It's breakfast. Eggs, pancakes, breakfast meat, coffee, juice. It might say more about the offended person if they need specific breakfast guardrails. The other person may have accidently filtered out a potentially life-changing disaster.
There are lousy and lazy ways to use LLMs. There are also enlightening and powerful ways. Find someone curious who will explore them. Don't date people who habitually fight anything new.
I think a lot of the problems we have now is an unbalance on this, an insufficiently strong conservative force. "Conservative" political movements stopped filling this role a couple generations ago in favor of a sort of radical revanchism. They want to change everything too, just in a different way.
People who fight every change will always lose. But they are still valuable to us.
It means you don't have a reason for rejecting the thing in question. Just crossing your arms with scowl and pretending things were better in years of yore.
Having a reason why you think the thing is bad, and should not be accepted is the better position to have.
There are some options that are clearly 99% superior than other options. But that 1% happens enough to create an extinction event for humanity if everyone relies on it.
People are going to oppose good ideas from their very core for no discernable rhyme or reason. But at the same time, it's not necessarily irrational - our genetics are hardcoded to hedge collectively in weird ways.
Sure. But date the intellectually curious. They are way more fun.
Wikipedia didn't start until 2001. There was no 90s Wikipedia criticism. It didn't exist.
I don’t mean to say that we should outsource our thoughts to LLMs. In this case it’s clear that even if they had just searched “places for xyz” on maps, it would’ve been just as bad - what they wanted was effort, shopping around + research with pros, cons, and discussions, not just “not an LLM”. Valid criticism!
Also, avoid dating people who assume that if you want to avoid the influence of genAI you "habitually fight anything new".
[1] by that I mean those that keep saying everyone around them they ask chatGPT or you should ask chatGPT for whatever is the subject of the conversation.
I think using shortcuts as a warning sign is okay but the author here uses it as a hard filter. She lacks curiosity. She doesn’t ask the groom why he chose that venue. She didn’t ask how he used ChatGPT. She just cared that he heard about it from ChatGPT rather than what… from an ad in a wedding magazine? A google search? Asking friends? Hiring a planner? That lack of curiosity makes her look like a crap journalist and drives her delusional shortcut, probably to her own detriment.
This is an author who simply does not understand how people actually use LLMs. She thinks it is a recommendation engine because her only use of it was to type Google searches into it. The chat interface encourages this sort of short question so I cannot really blame people who don't know you can enter enter tens of thousands of words.
I have colleagues who use it like this, and one who has proudly never used it at all. I politely congratulate him on this when he brings it up because I know it is very important to him for whatever reason, but it is meaningless, like someone proudly saying they haven't used Google Maps. He is apparently the only person I know who the author could date, and his ego as a lover of writing is so important to him that he eschews technology to protect himself.
"Find the most interesting, quirky, or just plain unusual attractions that locals are pleasantly surprised by, searching using {local language}"
The models have complex behavior surfaces and when they interact with external data, the dynamics pliable. You can get very much non-basic stuff for whatever you want.
But anyway, the article was about dating and I'd hope whomever I was dating had enough experiences to have their own opinions about at a minimum a cafe for breakfast.
Same logic applies to using Google, I guess? Or any mapping app? How about Yelp?
Not necessarily. The same logic would apply if you were just robotically using Google user ratings (e.g. "this is nearby and it has 5 stars, lets go there"), but it wouldn't if you looked through the results and used some taste, judgement, and experience.
Why do people make that assumption? That would be the dumbest possible way to use AI.
I don’t think it’s unfair to assume some non-trivial percentage of people use these tools in a very dumb way.
I hope we get some interesting psychological studies out of this sort of story in the coming years. Maybe we can learn a thing or two about ourselves.
Perhaps the man in question is a thoughtful and curious person who was utilizing all of the resources available to him to provide a great experience to someone he cared about.
If I went on a few dates with someone and was really having a great time, but they told me that all their planning and conversation pieces were from a gen AI platform, I absolutely wouldn't like that. If they used it to audit some code to check for any performance considerations though? Those are so different, at least to me.
This quote:
> A good friend recently told me that she spent a night with a man, and in the morning suggested they get breakfast together. He took out his phone, opened ChatGPT, and asked for restaurant suggestions. Why get close to someone who outsources decisions, including the fun ones like picking where to eat?
Is one I find fair. There is some humanity to choosing some place to eat with another human. It also doesn't feel crazy to me to outsource the gathering and ranking to ChatGPT, though.
Then a lot of the article banks on "societal harm":
> OK, so ChatGPT helps you write your grocery list. Does your individual convenience outweigh the societal harm it can cause?
With some of the things that the article points out to, like water and power usage, I can also see the argument there. But then you can point to so many other things with similar arguments. A classic being "you're using an iPhone right now, don't you know how those are made".
This quote also feels along these lines to me:
> Pereira thinks that using ChatGPT “shows such a laziness”.
>
>“It’s like you can’t think for yourself, and you have to rely on an app for that,” she said.
But we all use Google and a million other services that "remove thinking". This just feels like an additional step (depending on the case, as mentioned above).
I have no idea where to stand on any of this. There's so much to consider and I feel like I can only find one side or the other online to read about. No one with a middle ground that makes sense (that I've stumbled upon).
That said, I do think there's a huge difference between having it draft your love letters for you and having it tell you how to best treat a cut you got on your finger.
It's fairly telling that the people who hold these strong opinions think of themselves as strong independent thinkers but are slavishly devoted to the opinions of an algorithm.
I think there are a lot of decent reasons to be skeptical of AI. But this doesn't read like someone who is thoughtfully cautious and has concerns over actual issues with LLMs - this reads like someone in the 50s reading about microwaves and committing to never own one.
https://blogs.ubc.ca/etec540sept13/2013/09/29/socrates-writi...
I would bet that back in days some women would get ick from men not remembering everything but writing it down.
> ‘OK, so ChatGPT helps you write your grocery list. Does your individual convenience outweigh the societal harm it can cause?’
look at this garbage reasoning. "oh no a shopping list, IT IS THE END OF THE WORLD"
I wonder if she has such strong opinions on social media and short video content.
> I probably won’t find what I value in someone who thinks “productivity” means asking an app to summarize a movie plot so they don’t have to waste their time, you know, watching it.
Reasonable, but it would be very foolish to assume that everyone that uses LLMs uses it in such a stupid way. Everyone is free to have their own opinion, and my opinion is that the author didn't even try to go deep into the subject, just feeling the ick and reasoning it later.
> We know that the energy-intensive tech drains our water supply and hikes electricity bills.
Before GenAI, there was Google, AWS, Facebook, not to mention Netflix and friends. A single streaming episode probably uses more resources all told than reasonable ChatGPT use over a month.
> It is sold as a placebo for human connection; lonely, disconnected people finding companionship or even falling in love with code is not as much a sci-fi plot point as it is just the way things go now.
Anything/everything possible is sold as a placebo for human connection because loneliness is a powerful human motivator (Facebook, Instagram, TikTok, etc. etc.) Before that: TV.
> The megarich tech bros in charge of all this think in terms of profit first and people second.
Most profitable corps think about profit, yes. The idea is that they can't profit unless people get something good out of it. If they can profit without people getting something good, it's a failure in the system, so the correct target is market breakdown: bad regulation, unbroken monopolies, fraud, etc.
> OK, so ChatGPT helps you write your grocery list. Does your individual convenience outweigh the societal harm it can cause?
...
Let’s just say intelligence is changing and wisdom is in deep collapse using this tech.
Our drive to be creative and chance failure and seek expert help dissolves in a general tool that automates what it is to be human.