Does this mean microsofts potential billion dollar aquisition of openAI is a bad idea because the IP is already out there and other companies are catching up?
Meet “Claude”: Anthropic’s rival to ChatGPT
121–130 of 165 posts
Re: Meet “Claude”: Anthropic’s rival to ChatGPT
#122we need less censorious AIs not more ... the claim that it's somehow 'ethical' to have a guy baking in his opinions about things in a tool used globally is absurd to anyone who ever read anything about ethics
We already had a learning AI without hard limits. Remember Microsoft's Tay? Ah, right, it was shut within days because it quickly became an asshole. ( https://www.theverge.com/2016/3/24/11297050/tay-microsoft-ch... ) It's not even going to be objectively wrong a lot of the time. For example slaves are indeed a pretty efficient way to run a business. It's our limits and ethics that stops (most of) us from doing it. Wi…
> Remember Microsoft's Tay? Ah, right, it was shut within days because it quickly became an asshole.
If we consider the context, which is not something that posts on twitter under Microsoft's brand name, but something you communicate with in private. Who exactly are you worried about here, the person who will prospectively coerce the language model into being an asshole? If they don't want to do that, they could just not do that.
If I make ChatGPT say something egregious, and post that on Twitter or Facebook, I'm posting it, and I'm liable, just as I would be liable if I used a word processor with spell checking to make text and post it on Twitter or Facebook.
> Without ethics, you'll likely always end up with a 4chan-bot instead of whatever you intended.
If I ask it to explain quantum physics to me in the style of Donald Trump because it is funny, and it does it (as ChatGPT used to do), who exactly is being harmed and under what system of ethics, because as you may know, ethics is not objective or universal.
Re: Meet “Claude”: Anthropic’s rival to ChatGPT
#123> Claude feels not only safer but more fun than ChatGPT. I may be the minority here, but I really don't concern myself with ChatGPT safety and I am not entirely sure what the reason is why people are very worried about its safetly. It is safer than most things I have in my house, including a kettle, a saw, a hammer, a screwdriver, my actual PC, every kitchen appliance I have. Of course it can be misused, like any too…
"Safe" has gradually shifted in meaning because of the internet. Since interactions are limited to words and reactions it ended up generally meaning "don't upset anybody" for commercial and political reasons. For this reason ChatGPT refuses to write a fiction story in which somebody dies even though death is part of even some children stories. It hurts no one of course, but you can bet some people will be upset and start a controversy.
And in general this influence of the internet has reversed "actions speak louder than words". Textile companies support minorities with social media posts and printed tshirts while using factories with abused underage workers in underdeveloped countries. IT companies support the left side of politics while avoiding paying taxes. And so on. There's a huge and interesting discussion about this but is of in this case quite off-topic.
Re: Meet “Claude”: Anthropic’s rival to ChatGPT
#124> Claude feels not only safer but more fun than ChatGPT. I may be the minority here, but I really don't concern myself with ChatGPT safety and I am not entirely sure what the reason is why people are very worried about its safetly. It is safer than most things I have in my house, including a kettle, a saw, a hammer, a screwdriver, my actual PC, every kitchen appliance I have. Of course it can be misused, like any too…
I hope someone makes one of these things that has been trained without any concern for 'safety' or propriety, just for the sake of comparison.
There are already several models "in the wild" designed specifically to produce anime-styled "adult material". In Stable Diffusion for image generation, NSFW filtering is basically a post-processing step that can be disabled in a single line of code. I'm sure someone has already done it for Chat* models as well.
Re: Meet “Claude”: Anthropic’s rival to ChatGPT
#125Earlier quoted context omitted.
GPT3 works quite well at following instructions if adequately prompted, just not as well as ChatGPT. ChatGPT was trained separately for ability to follow instructions and for harmlessness (sic). Not only is a non-moralising ChatGPT possible, one was actually created during the research process. You may also wish to know that most readers of erotica are women.
> You may also wish to know that most readers of erotica are women. I know that, but that doesn't mean they want to get it anytime they prompt with their names. There's actually multiple anecdotes from OpenAI employees about this happening to them (with prompts like "write a diary entry about my day").
> with prompts like "write a diary entry about my day")
Although I don't know what sensible output anyone would hope from that prompt.
Re: Meet “Claude”: Anthropic’s rival to ChatGPT
#126Earlier quoted context omitted.
> You may also wish to know that most readers of erotica are women. I know that, but that doesn't mean they want to get it anytime they prompt with their names. There's actually multiple anecdotes from OpenAI employees about this happening to them (with prompts like "write a diary entry about my day").
> with prompts like "write a diary entry about my day") Although I don't know what sensible output anyone would hope from that prompt.
Don't have a diary on me, but seems to me if you fed it some examples it'd work fine on ChatGPT.
Re: Meet “Claude”: Anthropic’s rival to ChatGPT
#127> Claude feels not only safer but more fun than ChatGPT. I may be the minority here, but I really don't concern myself with ChatGPT safety and I am not entirely sure what the reason is why people are very worried about its safetly. It is safer than most things I have in my house, including a kettle, a saw, a hammer, a screwdriver, my actual PC, every kitchen appliance I have. Of course it can be misused, like any too…
"Safe" has gradually shifted in meaning because of the internet. Since interactions are limited to words and reactions it ended up generally meaning "don't upset anybody" for commercial and political reasons. For this reason ChatGPT refuses to write a fiction story in which somebody dies even though death is part of even some children stories. It hurts no one of course, but you can bet some people will be upset and s…
> It hurts no one of course, but you can bet some people will be upset and start a controversy.
The situation sucks, but sure, you are right. Still, I want a good tool, and I don't mind if someone holds me accountable if I misuse it, that is fine, but I want the tool, and I will get it, maybe not this year, maybe not next year, but in the next 5 years I'm sure I will have something as capable as ChatGPT is now without artificial restrictions available to me.
Just seems rather pointless that we have to play this silly safetyism game until then.
Re: Meet “Claude”: Anthropic’s rival to ChatGPT
#128Earlier quoted context omitted.
We already had a learning AI without hard limits. Remember Microsoft's Tay? Ah, right, it was shut within days because it quickly became an asshole. ( https://www.theverge.com/2016/3/24/11297050/tay-microsoft-ch... ) It's not even going to be objectively wrong a lot of the time. For example slaves are indeed a pretty efficient way to run a business. It's our limits and ethics that stops (most of) us from doing it. Wi…
> Remember Microsoft's Tay? Ah, right, it was shut within days because it quickly became an asshole. If we consider the context, which is not something that posts on twitter under Microsoft's brand name, but something you communicate with in private. Who exactly are you worried about here, the person who will prospectively coerce the language model into being an asshole? If they don't want to do that, they could just…
> If I make ChatGPT say something egregious, and post that on Twitter or Facebook, I'm posting it, and I'm liable, just as I would be liable if I used a word processor with spell checking to make text and post it on Twitter or Facebook.
While arguably that would be the case, practically it might be a different look for the company whose algorithm generated the text in the first place—and from the audience, which might think that it somehow represents the viewpoint of the company.
I mean surely you have noticed that social media can get stirred up for no fundamentally sound reasons.
Re: Meet “Claude”: Anthropic’s rival to ChatGPT
#129Earlier quoted context omitted.
The thing is used to write code ...
So is my editor, my computer, my fingers, my eyes, my brain, the internet, google, stackoverflow, wikipedia, English, etc.
Yeah, but there is still a human between the code and its execution.
I'm sure one of these days someone will hook GPT directly to a terminal.
Re: Meet “Claude”: Anthropic’s rival to ChatGPT
#130They say Claude is "more verbose", and claim this is a positive. I disagree. My biggest criticism of ChatGPT is that its answers are extraordinarily long and waffly. It sometimes reminds me of a scam artist trying to bamboozle me with words. I would much prefer short, concise, precise answers.
ChatGPT can be very confidently wrong in few words too. It's a very flexible system that way.
I started asking ChatGPT some rather technical questions about Australian drug laws. First I asked it what schedule common ADHD medications were on, and it answered me correctly (schedule 8). Then I asked it what schedule LSD was on, and it told me it wasn’t on any schedule, because it was an entirely legal drug in Australia. Uh, I hope it doesn’t some day tell someone that and they actually believe it, because they may be in for a very unpleasant experience if the police happen to encounter them acting on it