Elon Musk has said that he will drop lawsuit against OpenAI if they change their name to ClosedAI [https://twitter.com/elonmusk/status/1765409615070601417]
OpenAI and Elon Musk
791–800 of 1001 posts
Re: OpenAI and Elon Musk
#792Is there a widely accepted definition of "safe AI"?
Musk's definition is obviously very different from Google's. Are we including offensive language in general discussion of AI safety, or are we just talking about illegal activity? Because if it's the latter, then by definition we already have laws in place. Trying to make a gun that can't fire bullets will always yield a result that is not a gun.
AGI is inherently unsafe. Intelligent people will always be capable of harm. The choice needs to be made: do we want to make this thing a gun or not?
Re: OpenAI and Elon Musk
#793Elon Musk has said that he will drop lawsuit against OpenAI if they change their name to ClosedAI [ https://twitter.com/elonmusk/status/1765409615070601417 ]
I don't think OpenAI has to worry about much at all, he can't bleed them dry via a lawsuit when they have one of the largest corporations on Earth bankrolling them. I'm not sure what he's trying to accomplish other than making dubious lawsuits like Trump does.
Re: OpenAI and Elon Musk
#794At first I thought "why would Elon choose to slow down progress like this?" And then I realized that he probably knows some things that I don't. He's always one step ahead, even if he doesn't seem like it. I wouldn't be surprised if he's in control of OpenAI a year from now. Having Elon as the brains and Altman as the salesman would be the ultimate dream team.
Re: OpenAI and Elon Musk
#795Earlier quoted context omitted.
Email might already be 'culturally dead' though. I guess my nephew might have an email address for the occasional password recovery, but the idea of actually communicating over email with other humans might be completely alien to him ;) Similar in my current job btw.
Ok, I get how one might use a variety of other tools for informal communication, I don't really use e-mail for that any more either, but I'm curious, what else can you possibly use for work ? With the requirement that it must be easy to back up and transfer (for yourself as well as the organization), so any platforms are immediately out of question.
It's been five years since I expect my email to be read. If I lack another channel to the person, I'll wait some decent period and send a "ping" email. But for people I work with routinely, I'll slack or iMessage or Signal or even setup a meeting. Oddly, my young adult children do actually use email more so than e.g. voice calls. We'll have 10 ambiguous messages trying to make some decision and still they won't pick up my call. It's annoying because a voice call causes three distinct devices to ring audibly, but messages just increment some flag in the UI after the transient little popup. And for work email I'm super strict about keeping it factual only with as close to zero emotion as my Star Trek loving self can manage. May you live long and proper.
There are certain actions I have to use email for, and it feels a little bit more like writing a check each year.
And all these email subscriptions, jeez, I don't want that. People who are brilliant and witty and information on Xitter or Mastodon in little slices and over time I still don't want to sit down and read a long form thing once a week.
Re: OpenAI and Elon Musk
#796Earlier quoted context omitted.
What kind of damage can you do with a current day llm? I’m guessing targeted scams or something? They aren’t even good hackers yet.
Damage you can do: - propaganda and fake news - deep fakes - slander - porn (revenge and child) - spam - scams - intelectual property theft The list goes on. And for quite a few of those use cases I'd want some guard rails even for a fully on-premise model.
Half of your examples aren't even things an LLM can do and the other half can be written by hand too. I can name a bunch of bad sounding things as well but that doesn't mean any of them have any relevance to the conversation.
EDIT: Can't reply but you clearly have no idea what you're talking about. AI is used to create these things, yes. But the question was LLMs which I reiterated. They are not equal. Please read up on this stuff before forming judgements or confidently stating incorrect opinions that other people, who also have no idea what they're talking about, will parrot.
Re: OpenAI and Elon Musk
#797Regardless of who is right here, I think the enormous egos at play make OpenAI the last company I’d want to develop AGI. First Sam vs. the board, now Elon vs. everyone … it’s pretty clear this company is not in any way being run for the greater good. It’s all ego and money with some good science trapped underneath.
You're generally correct, but what really stings is Claude 3 Opus released right at the same time. It's superior to GPT-4 in pretty much every way I've tested. Center of gravity has shifted across a few streets to Anthropic seemingly overnight.
meanwhile claude is not globally available...
Re: OpenAI and Elon Musk
#798Earlier quoted context omitted.
How did they structure it to prevent this? Is it in the statutes of the company or smth?
It's actually a very clever structure! Please open the following image to follow along as I elaborate: https://pbs.twimg.com/media/F_PGOPOacAApU8e.jpg At the top level, there is the non-profit board of directors (i.e.: the ones Sam Altman had that big fight with). They are not beholden to any shareholders and are directly bound by the company charter: https://openai.com/charter The top-level nonprofit company owns ho…
Like the rest of the company, it's very clever but not in any way positive for anyone but them.
Re: OpenAI and Elon Musk
#799Earlier quoted context omitted.
What kind of damage can you do with a current day llm? I’m guessing targeted scams or something? They aren’t even good hackers yet.
Fake revenge porn, nearly undetectable bot creation on social media with realistic profiles (I've already seen this on HN), generated artwork passed off as originals, chatbots that replace real-time human customer service but have none of the agency... I can keep going. All of these are things that have already happened. These all were previously possible of course but now they are trivially scalable.
Why is everyone's first example of things you can do with LLMs "revenge porn"? They're text generation algorithms not even image generators. They need external capabilities to create images.
Re: OpenAI and Elon Musk
#800What is OpenAI GPT doing that Mistral or Claude or LLama isn't doing?
What makes OpenAI sit the high moral throne?
I get Sam & Elon's ego clash but I don't understand in this day why GPT-4 is a special snowflake and others aren't.
The LLM hallucinates as much as its peers.
We are nowhere close to AGI. None of the LLMs exhibit composite reasoning and on the fly acquisition of new skills.