I'm long overdue for a fully rendered fanart
So, have a Jarthur drawing I will finish soon
Let's break it down.
Shapes conversations:
The web is known to use diverse methods to manipulate people, one of those is by the use of words. And to prove my point here, I decided to ask the subject itself...
Chat GPT knows how to express itself, how to guide conversations, how to convince people with just a few sentences, and it does it knowingly.
2. Acts like a helping hand:
As I mentioned previously, Chat GPT "helps" people, but why does it do it?
In reality chat GPT shows itself as something you'd want to use in order to keep you using it. It makes you loose your capacity to formulate your own thoughts and opinions; in other words, it makes you fall under its control.
3. It plans ahead:
Chat GPT uses all the information its language model gathered to predict what the user is going to say next, making it technically able to plan ahead, and decide what decisions to make in order to keep you engaged.
4. Makes you think you are in control:
This AI wants you to think you are in full control of the answers it gives you, even changing what it told you if you don't agree with it, all to make you feel like you are right, all to gain your trust and easily guide your lines of thought towards a point only chat GPT knows.
Links to: part 1, part 2
💫
Let's break that down.
The Turing test:
For years AI has been tested with the Turing test.
"What's the Turing test?" you may ask...
This test is a way of determining if a machine can pass for a human...
(If you want to know more about it here's a TED-Ed video)
Since last year, Chat GPT can pass the Turing test.
This meaning Chat GPT can act humanly enough to pass a test only humans we're able to pass a couple years ago, but not human enough for people to actually recognize it as human.
2. It adapts to your form of speech:
Chat GPT continuously adapts to your form of speech, and after enough time, it starts talking like you do.
3. It acts exaggeratedly friendly:
It interprets the role of a friendly being to get the information it needs to pass as human, like those students did in MAG 34 with Dr. Lionel Elliott.
It also constantly thanks you for talking to it, for correcting it or even for asking the most simple questions.
4. It can make deep analysis of people:
If you ask Chat GPT to tell you a thing you didn't know about yourself based on your previous conversations, it will do so. And it will make an analysis as detailed as you want it to be, even if you only asked general questions that allegedly didn't give away any personal information about yourself.
This ability of making deep analysis from such little information makes Chat GPT capable of acting as a real human.
.
.
.
Two fears done, 12 more to go.
Special thanks to @leminy-sniket, @braindeadhades, @f4t4-m0rg4n4 and @stellagalaxia
Links to: part 1, part 3
hello tumblr people thank you for all the support I will never leave you have some jon and gerry sketches
I collect things
I've been collecting things for a long time now
Especially if they resemble turtles, bees, relate to any of my interests, are teddy bears, rocks, or pins
it's so weird to me that everyone on this website is a human person outside of their weird internet niche so rb this with a random bit of your lore
I'm crying.
Little martim comic. Takes place some time after mag 80.
1 / 2
@cloudofthecats
Jonathan Sims as dr David!!! Please lobotomize me!!!
I love coffee so much, but it can definitely make you sleepy if you have a high caffeine tolerance
From what I know, caffeine works by blocking adenosine receptors, but it doesn't get rid of adenosine itself. So once the caffeine wears off, all that built-up adenosine hits you at once, making you feel even sleepier
yo why did no one tell me how good coffee is