Paul Graham came out publicly to defend Sam, and we instantly have this blog post about, wait, just a sec, let's dissect actually why Sam is still evil.
Can we believe that Sam could actually be a good person? Today, Kara Swisher in her podcast on Pivot said, "Every time I tell people I actually like Sam, they become widely offended".
You’re applying pretty black and white moral values to a post that, at least to me, didn’t read that way at all. One can like Sam Altman as a person while wishing he was more transparent in some of his business dealings.
The post may not be at the extremes, but if the author has been following the issue as closely as it seems, they must be aware that there are people boldly proclaiming Sam Altman to be a sociopath on a daily basis.
The issue has become polarized, for reasons I don't rightly understand, but nevertheless this is where we have ended up.
To write on the topic in this environment it would be advisable to be clear on what they are saying, what they have an issue with, and what the appropriate remedy would be.
To just throw out some insinuations in a "I'm just asking questions" manner doesn't in-itself condemn a person. It isn't happening in isolation though. No snowflake believes itself to be responsible for the avalanche.
> To write on the topic in this environment it would be advisable to be clear on what they are saying, what they have an issue with, and what the appropriate remedy would be.
Its insane how far people are willing to project on their feelings towards Altman. Look at this quote in this thread.
"But it's worth noting that much of Sam Altman's presentation is just a mask (one he puts a ton of effort into and is good at maintaining), even if he's still less evil than the Sacklers or a mob boss."
> Can we believe that Sam could actually be a good person?
Depends on what good means to you. This is a person that we have evidence on repeatedly using these kind of underhanded techniques. Maybe he's not physically hurting anyone, but this is a person I would avoid.
Are you saying the public won't or shouldn't care? Altman wants public trust to say what regulations should and should not be made. Dishonesty is relevant.
No, if anything that’s a pretty fake controversy too.
Ricardo Montalban had a great quote about the life stages of an actor, enumerating them as follows:
1. Who is Ricardo Montalban?
2. Get me Ricardo Montalban.
3. Get me a Ricardo Montalban type.
4. Get me a young Ricardo Montalban.
5. Who is Ricardo Montalban?
As far as I can tell, Johansson’s complaint is that when OpenAI reached out to her for voice acting and she turned them down, that they instead got a Scarlett Johansson type, and that OpenAI should be categorically prohibited from hiring any voice actor who sounds like her at all. Which is not how acting has ever worked, but for some reason the topic of artificial intelligence gets a lot of people worked up to the point of artificial stupidity.
Midler v. Ford is more relevant legally than Ricardo Montalban's wit. In short deliberate mimicry is not allowed.
OpenAI's public claims about how they produced the Sky voice followed Johansson's public statement. They could be true or false. We don't know what claims or evidence they gave Johansson's counsel.
> As far as I can tell, Johansson’s complaint is that when OpenAI reached out to her for voice acting and she turned them down, that they instead got a Scarlett Johansson type
Agreed, and according to Midler v. Ford that is not permitted:
"We hold only that when a distinctive voice of a professional singer is widely known and is deliberately imitated in order to sell a product, the sellers have appropriated what is not theirs and have committed a tort in California."
I’ve never seen that standard applied to acting. Particularly voice acting. Midler v. Ford was a dispute between a singer and a company that used a sound-alike singer, singing a cover of a Bette Midler song in a commercial, to falsely imply a sponsorship that didn’t exist. Totally different case.
There’s a difference between impersonating someone with the explicit intention of falsely giving the impression they are involved in a project which they are not, and simply hiring an actor (or a voice actor) who can give a performance that’s similar or reminiscent of another.
> You've never seen it, huh? Google is super easy to use.
> from my perspective it seems quite obvious you're not arguing in good faith
I’m arguing in good faith, in the sense that I’m expressing my own genuine rationale for my own opinions. Your inability to cope with that and remain civil is the only show of bad faith in this entire discussion.
It’s quite a few things. I didn’t find his claims of ignorance around the non-competes, for example, particularly compelling.
But all of that is quite separate from these conflicts, which are entirely a matter for Altman and his investors, investors who have no reason to complain about him.
> What in the article is underhanded? Worst case, he has undisclosed conflicts of interest.
undisclosed = underhanded
> Does Altman have a Trump-like wake of ruined careers and lost riches among former allies? Everyone he's been close to seems to have done well from it.
I'm not talking about Trump, and I don't think Trump should be a reference for what is or isn't acceptable.
People are angry about the "Open"AI debacle and he's been publicly in favor of being very paternalistic and controlling (likely for his material benefit as much as safety). It's fair that he's taken some flak for those things, he's trying to control the direction of society at large and people want a say. I don't think he's evil, but I can see why people would perceive him as paternalistic or even a bit patronizing.
> Paul Graham came out publicly to defend Sam, and we instantly have this blog post about, wait, just a sec, let's dissect actually why Sam is still evil
I'm not seeing good and evil in this post. It's calling Sam out for not being transparent. Given he's elevated OpenAI, in public testimony, to an extinction-level threat to humanity, that lack of transparency is of public concern.
Not being transparent doesn't make him evil, doesn't mean he is unlikeable and doesn't per se mean he's dishonest. (Though OpenAI and he do have a likeability problem, at least in politics, albeit one I think they can fix.)
That’s exactly what I was going for — the issue of transparency here. It wasn’t dissecting why he’s “bad”, it’s that the public statements don’t match up with financial realities.
Maybe next time I could press more about the transparency factor, but I thought it was concise enough.
Obviously not, but as I state in the article he has plead:
> “He owns no stake in the ChatGPT developer, saying he doesn’t want the seductions of wealth to corrupt the safe development of artificial intelligence, and makes a yearly salary of just $65,000.”
According to OpenAI themselves.
So he takes a “low” salary and no ownership as to, according to him and the company, not influence his decisions in the pursuit of financial gain — yet that’s a complete omission of the whole truth.
I’ll stop short of calling it a flat-out lie, but a mischaracterization of reality for sure.
I'd love for Altman to explain why they decided to start Worldcoin in Africa, and by offering bigger and bigger signup incentives for their whole retinal scanning thing, to the point where at times it could be two month's wages for some people...
Keep in mind, when Swisher says she likes Sam, what she means is, to quote her Twitter: "Sam Altman is no different than most of the talented ones, which is to say, aggressive, sometimes imperious and yes, self-serving."
People don't divide cleanly into "good" and "evil" buckets, and CEOs in general tend to be ruthless deal -makers. But it's worth noting that much of Sam Altman's presentation is just a mask (one he puts a ton of effort into and is good at maintaining), even if he's still less evil than the Sacklers or a mob boss.
I mean, yeah, same thought after seeing the signatories. What are some of the cliches being used around here ? Toothpaste is out of the tub? Arrow has left the bow. The dye is cast. The ship has sailed. (Thanks ChatGPT).
The pee is in the pool. The black swan has left the barn.
And yeah, I had a laugh at the signatories. Of course my heart goes out to the non-billionaires that might be out of a job. Or maybe us lucky duckies are going to travel the world on our new basic income trust funds?
Go take a look at the content of Civitai. Take everything you see there, and imagine what happens if you start prompting it with words that indicate things which may not be legal for you to see images of.
Please show me viable harm of GPT-4 that is higher than the potential harm from open sourced image generators with really good fine tuning. I'll wait, most likely forever.
Not sure what you mean, but for example, 2 separate competitors to DALL-E was released within months (SD and MJ). Arguable that both of these have since surpassed DALL-E's capabilities/ecosystem.
LLMs take vastly more resources to train and run than image generators. You can do quite a bit with SD on a few year old 4GB laptop GPU (that’s what I use mostly, though I’ve set up an instance with a better GPU on Compute Engine that I can fire up, too.)
GPT-NeoX-20B – an open (as in Open Source, not OpenAI) LLM intended as a start to move toward competing with GPT-3 (but still well behind, and smaller) requires a minimum 42GB of VRAM and 40GB system RAM to run for inference. The resources times time cost for training LLMs is…immense. The hardware cost alone of trying to catch up to ChatGPT is enormous, and unless a radical new approach that provides good results and insanely lower resource requirements is found, you aren’t going to have an SD-like community pushing things forward.
Will there be competition for ChatGPT? Yes, probably, but don’t expect it to look like the competition for Dall-E.
Interests align - those who keep the secrets are better off for keeping them (not just for punishment of breaking NDAs, but the fruits (pun intended) of revealing something wonderful).
> Similar to balancing other social issues, I don't believe private companies should make all of the decisions on their own.
Then give up majority control.
This article needs a listen in compassion. It's all, "no we're not", instead of, "we understand why people feel this way and this is what we need to do"
I think one of the comments said vulgar and predatory. I second this. You don't need a bootcamp to break into Data Science, or Data Engineering, or Data anything.
This is my main grip with YC touting startups as a great place to join, but not helping them negotiating fair equity. Yes, founders deserve credit because they took risk, but they need to spread the wealth for execution. It should almost be illegal to give employees such low equity, relative to what founders get.
Can we believe that Sam could actually be a good person? Today, Kara Swisher in her podcast on Pivot said, "Every time I tell people I actually like Sam, they become widely offended".