| Страници во темата: [1 2] > | Poll: Have you ever heard of the 'Dunning-Kruger effect' in the context of AI? Лице кое објавува дискусија: ProZ.com Staff
|
|---|
This forum topic is for the discussion of the poll question "Have you ever heard of the 'Dunning-Kruger effect' in the context of AI?".
View the poll results »
| | | |
What exactly do you mean in context of AI?
Like I watched astronaut AI video, and now I'm over confident considering myself as an astronaut?
Or you are saying that our translation rates are too high considering the AI cheap version?
Or you are saying that some AI engines are overconfident with their prices considering the free AI engines?
| | | | Samuel Murray 
Холандија
Local time: 04:56
Член (2006)
англиски на африканс
+ ...
ProZ.com Staff wrote:
"Have you ever heard of the 'Dunning-Kruger effect' in the context of AI?"
I assume this relates to how good people are at recognising hallucination. I'm guessing most people would tend to be overconfident about their ability to detect incorrect or implausible information, especially if the bot is being very convincing. I must suffer from this as well: I judge how likely it is that what the bot is saying is true based on a gut feeling and nothing more (unless I then Google for it).
| | | | | Sounds like you're trying to clarify whether the Dunning-Kruger effect, in the context of AI, refers | Apr 20, 2025 |
Gjorgji Apostolovski wrote:
What exactly do you mean in context of AI?
Like I watched astronaut AI video, and now I'm over confident considering myself as an astronaut?
Or you are saying that our translation rates are too high considering the AI cheap version?
Or you are saying that some AI engines are overconfident with their prices considering the free AI engines?
Sounds like you're trying to clarify whether the Dunning-Kruger effect, in the context of AI, refers to users becoming overconfident after limited exposure (like watching one flashy demo), or if it’s about how AI tools themselves are priced or perceived compared to their actual capabilities.
All three interpretations you mentioned make sense in different ways — it could be about user overconfidence, inflated expectations of AI's capabilities, or even pricing models that don’t match the tool’s actual value. It's definitely an interesting angle to explore in the context of language services and translation tools.
| | |
|
|
|
I’ve seen a few exemples of the Dunning-Kruger Effect in action well before the AI crazy hype…
| | | | | Everyone is overconfident | Apr 20, 2025 |
Samuel Murray wrote:
ProZ.com Staff wrote:
"Have you ever heard of the 'Dunning-Kruger effect' in the context of AI?"
I assume this relates to how good people are at recognising hallucination. I'm guessing most people would tend to be overconfident about their ability to detect incorrect or implausible information, especially if the bot is being very convincing. I must suffer from this as well: I judge how likely it is that what the bot is saying is true based on a gut feeling and nothing more (unless I then Google for it).
Everyone on this world sometimes is overconfident about certain skills, from kids to grown ups. There's no such thing as perfection, or in our case 'perfect translation'.
On every translation I and you can find 'mistakes' and replace them with synonyms and call our translations as true. You can make problems when there none. I don't know how this condition is called, but the main reason would be to one prove self as perfect or better than the other one, and if there aren't any errors he/she will making things up.
Anyway, if you don't try something, you would never know if you are good at it.
And if you want to know my opinion or my confidence in AI, look on this forum there are some posts from me about AI, Google Translate, Ace Translate and how all began.
[Edited at 2025-04-20 10:39 GMT]
| | | |
[quote]Samuel Murray wrote:
I assume this relates to how good people are at recognising hallucination. I'm guessing most people would tend to be overconfident about their ability to detect incorrect or implausible information, especially if the bot is being very convincing
I watched a video with a girl talking about an exam to which she came unprepared (what a surprise!). She somehow managed to use Chatty for all her questions and then let her friend copy the answers. The number of correct answers was zero. And then there's my fiendishly skeptical self. I watched a lengthy review of what AI can do, and the first conclusion they made was exactly what y thought when AI first came about: this thing is lying to you; re-check everything it says. And that's just what I do whenever I use AI to translate from languages I don't know (which is the only use I have for it). The second conclusion was that we should use AI for entertainment. No, sir. I don't want a bot to entertain me. I'm a self-entertaining unit
| | | |
Samuel Murray wrote:
ProZ.com Staff wrote:
"Have you ever heard of the 'Dunning-Kruger effect' in the context of AI?"
I assume this relates to how good people are at recognising hallucination.
..... I assumed the same, but perhaps in so doing, we're both showing DK in action
On reflection, I am increasingly uncertain of my ability to judge what a poll is asking.
| | |
|
|
|
Daryo
Local time: 03:56
српски на англиски
+ ...
Maria Teresa Borges de Almeida wrote:
I’ve seen a few exemples of the Dunning-Kruger Effect in action well before the AI crazy hype…
I've also seen quite a few, starting from the time of mechanical typewriters and heavyweight paper dictionaries, all the way up to now!
But you can't stop progress - Augmented Ignorance has brought a whole new order of magnitude to the Dunning-Kruger Effect, making it part and parcel of its modus operandi.
If AI generated Kudoz answers are any indication, spouting nonsense with maximum confidence seems to be today’s AI default mode, so if you've never ''heard of the 'Dunning-Kruger effect' in the context of AI'', it can be only be because you refuse to see it.
| | | |
Daryo wrote:
Maria Teresa Borges de Almeida wrote:
I’ve seen a few exemples of the Dunning-Kruger Effect in action well before the AI crazy hype…
I've also seen quite a few, starting from the time of mechanical typewriters and heavyweight paper dictionaries, all the way up to now!
But you can't stop progress - Augmented Ignorance has brought a whole new order of magnitude to the Dunning-Kruger Effect, making it part and parcel of its modus operandi.
If AI generated Kudoz answers are any indication, spouting nonsense with maximum confidence seems to be today’s AI default mode, so if you've never ''heard of the 'Dunning-Kruger effect' in the context of AI'', it can be only be because you refuse to see it.
If I refused to see it, I wouldn’t refer to AI as the AI crazy hype, would I? Anyway, the AI hype is just a massive example of the Dunning-Kruger Effect…
| | | | | Dunning-Kruger effect | Apr 20, 2025 |
Karl Wirk wrote:
Gjorgji Apostolovski wrote:
What exactly do you mean in context of AI?
Like I watched astronaut AI video, and now I'm over confident considering myself as an astronaut?
Or you are saying that our translation rates are too high considering the AI cheap version?
Or you are saying that some AI engines are overconfident with their prices considering the free AI engines?
Sounds like you're trying to clarify whether the Dunning-Kruger effect, in the context of AI, refers to users becoming overconfident after limited exposure (like watching one flashy demo), or if it’s about how AI tools themselves are priced or perceived compared to their actual capabilities.
All three interpretations you mentioned make sense in different ways — it could be about user overconfidence, inflated expectations of AI's capabilities, or even pricing models that don’t match the tool’s actual value. It's definitely an interesting angle to explore in the context of language services and translation tools.
Well, the question isn't clear to me, and I just wanted to clarify.
And yes, when I was a kid, I watched aggressive rollerblading videos, and I was overconfident and I started to blade. I wasn't a rollerblader, but at that time, as a teenager I thought I was on top of the world.
Anyway, these days we are hearing marketing of AI, like this model is the best, no, this model is the best, this 2.0 is now the best, oh, wait now there's another one which is best, my model can do this and that...
For only this price... For free...
This marketing in a way is the problem for one's overconfidence in AIs.
And that's why one's work should always be checked by a reviewer.
I check yours, you check mine, and everything is all good.
| | | | Siobhán Fulton
Соединети Американски Држави
Local time: 22:56
германски на англиски
| Augmented ignorance | Apr 20, 2025 |
Daryo wrote:
Augmented Ignorance has brought a whole new order of magnitude to the Dunning-Kruger Effect,
Love it!
| | |
|
|
|
| Poor vs good performers | Apr 20, 2025 |
Poor performers would chose to follow the commanding AI voice. But this is the case of the voice, of individual psychological features of a given performer, not the source
The topic rather relates to the ancient art of rhetoric
[Редактировалось 2025-04-20 17:31 GMT]
| | | |
Daryo wrote:
Maria Teresa Borges de Almeida wrote:
I’ve seen a few exemples of the Dunning-Kruger Effect in action well before the AI crazy hype…
I've also seen quite a few, starting from the time of mechanical typewriters and heavyweight paper dictionaries, all the way up to now!
But you can't stop progress - Augmented Ignorance has brought a whole new order of magnitude to the Dunning-Kruger Effect, making it part and parcel of its modus operandi.
If AI generated Kudoz answers are any indication, spouting nonsense with maximum confidence seems to be today’s AI default mode, so if you've never ''heard of the 'Dunning-Kruger effect' in the context of AI'', it can be only be because you refuse to see it.
I agree to "ignorance". It is precisely what this so-called "intelligence" is. Knowledge implies understanding, which a computer program does not do.
But the branding IS clever. It justifies the gradual spoliation of yet another (relatively) free and common resource, to the profit of a new ruling minority. With much cynicism, as usual.
| | | | | Inverse effect | Apr 21, 2025 |
Charlie Bavington wrote:
On reflection, I am increasingly uncertain of my ability to judge what a poll is asking.
Me too. This does remind me, though, of a related phenomenon, a kind of inversion of the Dunning-Kruger effect.
Some people, when considering some skilled activity that they don't know how to do, will assume it must be very difficult ("I could never learn to paint portraits in a million years").
Other people will assume it must be very easy ("If I had a bit more free time I could be a brilliant portrait artist/concert pianist/brain surgeon/tennis player").
I'm sure we've all met people of both kinds. I know one individual from the second category who thinks that ChatGPT is utterly super. He recognises that the AI bot produces hallucinations from time to time, but he clearly believes he can distinguish them without difficulty.
He was also very pleased when ChatGPT generated a highly positive assessment of his own character and abilities.
So I'm wondering if there's a connection between the two personality traits.
| | | | | Страници во темата: [1 2] > | To report site rules violations or get help, contact a site moderator: You can also contact site staff by submitting a support request » Poll: Have you ever heard of the 'Dunning-Kruger effect' in the context of AI? | Protemos translation business management system | Create your account in minutes, and start working! 3-month trial for agencies, and free for freelancers!
The system lets you keep client/vendor database, with contacts and rates, manage projects and assign jobs to vendors, issue invoices, track payments, store and manage project files, generate business reports on turnover profit per client/manager etc.
More info » |
| | Draftsmith | Edit Translations Faster With Affordable AI You Can Trust
Draftsmith is an affordable AI editing tool for busy translators. It works directly in Microsoft Word, supporting the way you work
ProZ members receive a 20% discount on a single user subscription of Draftsmith.
More info » |
|
| | | | X Sign in to your ProZ.com account... | | | | | |
Login to reply/comment 
