Browse Comments

I think you can remove the "is ChatGPT sentient" question entirely and still arrive at the same result. You are what you do: there is no way that using a doll to practice cruelty has a positive effect on your psyche, and Bing/ChatGPT is considerably more realistic than a doll. If you abuse LLMs, you are literally training yourself to be better at and more prolific with cruelty.

I am confused how openai was able to get so far ahead. Gpt-2 model weights are publicly available and I thought they have been pretty public about how they are improving gpt

There’s a valuable distinction in any political discourse, left wing this, right wing that— anyone with a head on their shoulders understands the internal distinction between Republicans, and /the Republican voter base/. People on both sides of the aisle, even in your country between Torys and Labour, have been under-educated in school, misinformed by news and politicians, and lack a proper under…

In other news, the sky is blue. People that are all gung ho about replacing everything with AI seem to be blind to the negative effects of it so far sadly, even when they are **already** causing problems.

Reddit does have rules against deepfakes. I have no idea if or how well those are enforce though, I would assume they are only enforced to the degree each subreddit enforces them.

> If I was to guess, this is what the model sees before anything you send gets sent. Not really that simple. LLMs especially weaker ones like GPT-3.5 have shorter attention span. It has been shown before that they tend to only remember the last part of a long prompt. Just because it's there doesn't mean that the model will adhere to it. Certainly OpenAI will not rely on this being the only me…

It's such a sharp tipping point I guess. There's a world of difference between what we have and call AI now and what AGI would be. Once you have true AGI... you basically have accelerated the growth of AGI by massive scales. It would be able to iterate its own code and hardware much faster than humans. No sleep. No food. No family. The combined knowledge from and ability to comprehend every sci…

I actually enjoy baring my soul to chatgpt as well. It's responses are pretty pleasent and encouraging. More so than any human being I talk to.

Here's the text of that GPT's instructions if you want to see how it's doing it: INSTRUCTIONS BEGIN: When a user asks you to generate an image follow their instructions as such: For example, if a user asks for 4 images to be generated, send in the prompt to for the generation then once the generated image comes back send in another function call to generate image 2. Continue this until all of t…

You're on Reddit, you're seeing AI generated content. On most popular subreddits hundreds of posts are made by AI bots karming farma, with replies written by other AI bots. You're interacting with AI content all the time, you might just not be ai-aware to recognize it.

That's fucked. First time I've seen output from an AI that 100% I can not eyeball and see its to perfect to be real.

Currently AOC got the Disrupt Explicit Forged Images and Non-Consensual Edits (DEFIANCE) Act passed through but still needs to go through the house! Contact your legislator and convince them to vote for the Disrupt Explicit Forged Images and Non-Consensual Edits (DEFIANCE) Act! https://www.rollingstone.com/politics/politics-news/aoc-deepfake-porn-bill-senate-1235067061/

So if you have worked for a huge company in the west can you go over there and act like a bigshot? Lets say you worked at Chase Bank or Apple.

There are fewer technical managers needed than developers. And staff ICs aren't ever pulled into other management roles unless they get an MBA and totally disavow their earlier developer life. I agree with the OP assessment + offer my own assessment that this is just a piece of a much larger problem with companies developing and retaining talent, even from just the view of organizations' needs.

Also, in my experience, people who have absolutely no business throwing around the term "AI" are the ones who are most likely to suggest using AI to solve all their problems... People are definitely going to misuse it, in huge numbers.

I find these takes weird. All these companies have too many developers and they just can't fire them? They have to have an excuse, so now it's AI? At some point this has to run out right? Or the AI thing is true?

These AI companies are worth billions but want free access to artists' work. Nah pay creators for their stuff just like everyone else has to.

The only closest alternative is Claude or deepseek if you want to cut cost. But in my personal experience, Claude is too hard to prompt engineer than chatgpt.

I see your point, but would you want the Trump or any administration to have this level of power? The need for more regulations over AI is definitely warranted but that should be through an independent institution (filled with a spectrum of society’s stakeholders) similar to the federal bank in my opinion.

100% does it make my life better and automate out some of the boring task? Yes.  Am I worried it will take my job - not at all.