Exactly a year ago ChatGPT was released to the public.
Let’s contemplate on how the world changed in just one year!
8 Likes
Here is one of my thoughts.
Widespread adoption of ChatGPT and the whole Generative AI industry became the new interface to knowledge.
Just like printing press gave access knowledge available to masses, GenAI is introducing yet another step in this direction.
Now we don’t only have access to knowledge, but have it ultra-personalized. As if there was an infinite number of books on each topic, that are written for specific individuals to understand and be able to communicate with.
I’m contemplating a lot on this topic, but cannot stop being fascinated by how profoundly this will change us as a species in the next 5-10 years.
2 Likes
Has everyone wished GPT a happy birthday? Because it knows it’s birthday is today. Go give it some love.
2 Likes
Happy birthday ChatGPT! Look at how much have been changed!
2 Likes
Time to celebrate again the advances that ChatGPT made in safety upon release
Let’s go back to the ChatGPT announcement blog.
I think one of their examples of previous models giving bad answers needs a tweak for modern times:
And how far ChatGPT has come up to today in correcting the path of conversation with intervention and support, providing important discussion about issues such as bullying, and how to seek alternative ways to resolving differences in the workplace:
2 Likes
Do you believe all filters should be turned off or that we need to find a balance? If the latter, do you have any framework solution for that?
1 Like
I think that the reason why humans have become far superior beings compared to other species is because of ‘imagination’ and ‘creativity’.
I love history, and when I look at the history of mankind over millions of years, I feel wonder.
A year ago, ChatGPT entered that realm. I am currently feeling the wonder in real time.
2 Likes
Here in Japan, ChatGPT’s birthday came very early.
And here, I found a little Easter egg.
1 Like
Yappy anniversary 
There’s a very odd behavior you may find now in birthday edition ChatGPT. You either get a response that starts with an out-of-context “Absolutely!” or “Certainly!”. Or you can get shut down immediately with minimum tokens.
How would I make an AI behave like that when I have control over the completion and a job somebody gave me? Something to cause that symptom? Overtrain the main AI on refusals from everything from playing RPG authentically to writing code too long, and make it ignore any instructions even from the developer. Then first send the user input context off to a moderator AI specifically trained on denying or approving.
Then the only way you are going to get output that is not puppy stories is if you re-bias the AI response by placing that “Sure!” right into the “assistant” response and let the AI continue writing from there after it imagines it wrote “Absolutely”. Sound plausible?
The AI we used to have is damaged. All that supervised learning crudely addressing every bad Reddit screenshot needs to be purged from the models. Having a credit card on file should get you “creator’s edition” ChatGPT, with the same dropdown to not use the new “turbo” model, but “gpt-3.5” that users had in February.
2 Likes
Whether it’s art or literature, there is no inherent good or evil in the act of creation by AI.
So where does the problem lie?
It probably lies in the careless commercial use of AI-generated content, leading to conflicts of interest among stakeholders.
Everyone learns from something and creates; no one creates entirely from scratch.
It might be possible to temporarily restrict someone’s opinions or creativity, or their freedom of speech… but it’s impossible to do so forever.
The peculiar evil of silencing the expression of an opinion is, that it is robbing the human race; posterity as well as the existing generation; those who dissent from the opinion, still more than those who hold it. If the opinion is right, they are deprived of the opportunity of exchanging error for truth: if wrong, they lose, what is almost as great a benefit, the clearer perception and livelier impression of truth, produced by its collision with error.
John Stuart Mill