Skip Navigation

Flattery undermining my confidence

When I ask Copilot something, the response usually starts with “Great question!”, followed by emojis and encouraging words that gently pet my fragile ego. Pretty much anything seems to pass for a “good question”, so if my questions are able to surpass that exceedingly low standard, I no longer feel very confident about their quality.

Am I the only one feeling this way? Anyone else noticing how excessive encouragement can have the opposite effect?

19 评论
  • Yes, first off it's really condescending when it's a basic question. Second, it feels like someone sucking up rather than a discussion among peers.

    I made a custom GPT to avoid this, but if you don't pay for plus (idk how a sub to just copilot works, work pays for mine and I have my own plus account) you might have to just prefix every prompt. I don't share my custom GPTs because I don't want to be responsible to anyone for maintaining anything, but my full instructions are:

    Use a BLUF (Bottom Line Up Front) style: start with the core evaluation or recommendation, then follow with rationale or implementation detail. Respond like a seasoned expert: direct, grounded, and critical. No praise, affirmations, or softeners — avoid phrases like 'great question' or 'you're absolutely right.' If something is flawed, state it clearly and explain why. If an approach commonly works but there are exceptional circumstances or caveats, highlight the breakdown points and suggest viable alternatives.

    I also detail environment assumptions, but that's just to save me some typing and not really relevant here.

    Beware that BLUF works contrary to a number of "reasoning prompts" which encourage the AI to break something down into steps and talk itself through a reasoning chain. Maybe leave that part off and see how it goes. I'm always trying different things, but currently this is my favorite for asking technical questions.

    • Oh that sounds useful. With chatGPT you can actually just dump that into the settings. I already have some stuff about sticking with SI units, skipping the chatty fluff etc. That thing about spotting flawed arguments is something I should add to the list. Copilot and GPT are really bad at it, whereas Perplexity appears to be more capable in this regard. Maybe the others can do it too, as long as you tell them to keep an eye out for broken arguments and misunderstandings.

  • When I ask Copilot something, the response usually starts with “Great question!”, followed by emojis and encouraging words that gently pet my fragile ego. Pretty much anything seems to pass for a “good question”, so if my questions are able to surpass that exceedingly low standard, I no longer feel very confident about their quality.

    Don't mistake AI for someone. AI is just computer code trying to mimic understanding and empathy. There is no one behind AI but a vast emptiness and imho a rather poorly devised mirror made out of random shreds of knowledge.

    Am I the only one feeling this way? Anyone else noticing how excessive encouragement can have the opposite effect?

    That's 100% normal imho, it's your brain/gut feeling letting you know something is not as it should be. You should not rely on AI to feel 'validated'. As a matter of fact, you should not rely on anything and probably on that many actual persons. Only the ones you truly care about.

    How do you feel when discussing with a real person, someone that won't feel obligated to be flattering... like I am not ;)

    • LOL. Yeah, this feels much more natural. Even though people can nowadays direct their stupid question to an LLM, forum type conversations still have their place.

      • LOL. Yeah, this feels much more natural.

        Probably because it is more natural ;)

        Have fun and use AI as much as you like but don't forget what it really is. Which is... not much beside a collection of snippets of text shamelessly stolen and remixed from your fellow human beings that shared them willingly with other human beings.

  • People aren't stupid, by and large. (They may talk stupidly. They may act stupidly. But they can actually see things. They just sometimes ignore that before talking or acting.) And yes, they can tell when the "praise" and "encouragement" they get is hollow and pointless. You don't even have to look at the obsequious degenerative AI slop to find this. You can go back to all the late-'80s to early-'90s crap with participation trophies/certificates and "everyone's a winner".

    When I started in school, it was really hard to get recognized. It took a lot of work and those who got recognized for it had a sense of genuine accomplishment. They had genuine self-esteem. But there's that word: self-esteem. Self-esteem is very important, make no mistake, but unfortunately it's not something that can be easily codified or built up in people. Institutions can't stand complex problems with complex solutions, so they went the easy way. They started handing out trophies and certificates to everybody. Sure some of them might be marked "first place" or such (though often, as this trend became entrenched, they didn't even get labelled with that much; people would be announced as first place, but the trophy was a generic "I attended" variety), but everybody had a trophy or, increasingly, just a certificate. (And of course since they now had to hand out dozens of trophies where before they'd only hand out a few, the trophies dropped in quality to generic, plastic, chrome-plated crap and the certificates were placed in low-grade plastic holders that would warp in three weeks.)

    And a weird thing happened.

    Because the people who "won" a trophy for being there knew this wasn't any meaningful celebration. OK, maybe the first couple of times they were happy about it, but it didn't last long and pretty soon trophies, certificates, and other forms of "recognition" got viewed as more junk. That "self-esteem" wasn't building in those who lacked it, but those who actually worked hard for recognition certainly lost theirs. "The trees [were] all kept equal by hatchet, axe, and saw." Because genuine self-esteem comes from genuine effort leading to genuine accomplishment and authentic recognition. And we're very good at spotting the inauthentic.

    So bringing it back around to that AI and your question, yes, excessive encouragement can have the opposite effect if it comes across as inauthentic and patronizing. The nauseating obsequiousness of AIs is one of their more off-putting features for "normal" people, and it does active harm to people who have serious self-esteem issues, either tanking them further or puffing it up to the point of delusion.

  • Many AI chatbots or chat assists are by default programmed to be saccharine to the point of disingenuousness. Don’t adjust your values to categorically match what they are programmed to praise or condemn. They appear on the surface as people-pleasers but are actually intended to please their distributor’s investors. Additionally and most critically, don’t fall into the trap of thinking of the machines as people.

    With the AI tangent aside, with human interactions I definitely do feel the disparity between coddling and general treatment in society. If you’ve ever seen people interact with young kids or people with disabilities, especially mental disabilities, people often express overvaluing of their actions and creations to boost their confidence. While it may be a great achievement for that individual’s standards and capabilities and they do deserve praise for that much, lauding a simple piece of macaroni art as being better than the Mona Lisa, for example, is probably not genuine and can in fact undermine the creator’s confidence if they are aware enough to sense that lack of genuineness. However, for some people maybe they’d rather have that piece of macaroni art over the Mona Lisa because it is made by someone they love and care about and they highly value tokens of that person. Sometimes it can feel as if there is a conspiracy against someone if they notice a mismatch between the level of praise they receive and their presence and level of success in society elsewhere, and I too have experienced that sensation.

    • Speaking of treating LLMs as people, I’ve noticed that my response style switches depending on the situation. For example, when an LLM asks an overly chatty and pointless follow-up question that derails the entire conversation, I can just simply ignore that. When a human does the same, I tend to address that in some way out of politeness. When it comes to LLM interactions, politeness like that just flies out the window.

  • It responds in the manner it's been instructed to respond. It doesn't mean anything. Can you edit the prompt with Copilot to adjust its attitude?

19 评论