Live data from Hacker News

OpenAI departures: Why can’t former employees talk?

vox.com

371–380 of 1001 posts

Re: OpenAI departures: Why can’t former employees talk?

#371
post #319

Earlier quoted context omitted.

Agreements with surprise terms that only get detailed later tend not to be very legal.

How do you know there isn't a very clear term in the employment agreement stating that upon termination you'll be asked to sign an NDA on these terms?

Why not just get it signed then? Your signing to agree to sign later?

Re: OpenAI departures: Why can’t former employees talk?

#372
post #354

Earlier quoted context omitted.

Clever, but no. The argument about LLMs not being copyright laundromats making sense hinges the scale and non-specificity of training. There's a difference between "LLM reproduced this piece of copyrighted work because it memorized it from being fed literally half the internet ", vs. "LLM was intentionally trained to specifically reproduce variants of this particular work". Whatever one's stances on the former case,…

Which has been established in court where?

And it matters how? I didn't say the argument is correct or approved by court, or that I even support it. I'm saying what the argument, which OP referenced, is about, and how it differs from their proposal.

Re: OpenAI departures: Why can’t former employees talk?

#373

Earlier quoted context omitted.

Alignment in the realm of AGI is not about getting everyone to agree. It's about whether or not the AGI is aligned to the goal you've given it. The paperclip AGI example is often used, you tell the AGI "Optimize the production of paperclips" and the AGI started blending people to extract iron from their blood to produce more paperclips. Humans are used to ordering around other humans who would bring common sense and…

Right and that's why it's a farce. > Whoa whoa whoa, we can't let just anyone run these models. Only large corporations who will use them to addict children to their phones and give them eating disorders and suicidal ideation, while radicalizing adults and tearing apart society using the vast profiles they've collected on everyone through their global panopticon, all in the name of making people unhappy so that it's…

Don't worry, certain governments will be able to use these models to help them commit genocides too. But only the good countries!

Re: OpenAI departures: Why can’t former employees talk?

#374
post #319

Earlier quoted context omitted.

Agreements with surprise terms that only get detailed later tend not to be very legal.

How do you know there isn't a very clear term in the employment agreement stating that upon termination you'll be asked to sign an NDA on these terms?

One particularly sus term in my employment agreement is that I adhere to all corporate policies. Guess how many of those there are, how often they're updated, and if I've ever read them!

Re: OpenAI departures: Why can’t former employees talk?

#375

It's time to find a lawyer. I'm not one but there's an intersection with California SB 331, also known as “The Silenced No More Act”. while it is focused more on sexual harrasment, it's not limited to that, and these contracts may run afoul of that. https://silencednomore.org/the-silenced-no-more-act

No it’s either a violation of the NLRB rule against severance agreements conditioned on non-disparagement or it’s a violation of the common law rule requiring consideration for amendments to service contracts.

Re: OpenAI departures: Why can’t former employees talk?

#376

Earlier quoted context omitted.

NDAs don’t touch the copyright of your speech / written works you produce after leaving, they just make it breach of contract to distribute those words.

Following the legal defense of these companies, the employees wouldn’t be distributing any words. They’re distributing a model.

They’re disseminating the information. Form isn’t as important as it is for copyright.

Re: OpenAI departures: Why can’t former employees talk?

#377
post #175

Earlier quoted context omitted.

> Superintelligence will be the most impactful technology humanity has ever invented, and could help us solve many of the world’s most important problems. But the vast power of superintelligence could also be very dangerous, and could lead to the disempowerment of humanity or even human extinction. Superintelligence that can be always ensured to have the same values and ethics as current humans, is not a superintelli…

You might be interested in how CEV, one framework proposed for superalignment, addresses that concern: https://en.wikipedia.org/wiki/Friendly_artificial_intelligen... > our coherent extrapolated volition is "our wish if we knew more, thought faster, were more the people we wished we were, had grown up farther together; where the extrapolation converges rather than diverges, where our wishes cohere rather than interfe…

Is there an insightful summary of this proposal? The whole paper looks like 38 pages of non-rigorous prose with no clear procedure and already “aligned” LLMs will likely fail to analyze it.

Forced myself through some parts of it and all I can get is people don’t know what they want so it would be nice to build an oracle. Yeah, I guess.

Re: OpenAI departures: Why can’t former employees talk?

#378

Earlier quoted context omitted.

Exactly. There’s no ethical way to gain ownership of a billion dollars (there’s likely some dollar threshold way less than 1B where p(ethical_gains) can be approximated to 0) A lot of people got screwed along the way

i think a lot of people have been able to become billionaires simply by building something that was initially significantly undervalued and then became very highly valued, no 'screwing'. there is such thing as a win-win and frankly these win-wins account for most albeit not all value creation in the world. you do not have to screw other people to get rich. whether people should be able to hold on to that billion is a…

I wouldn't know, I'm not a billionaire. But when you hear about Amazon warehouse workers peeing into bottles because they they don't have long enough bathroom breaks, or Walmart workers not having healthcare because they're intentionally scheduled for 39.5 hours, it's hard to see that anyone could get to a billion without screwing someone over. But like I said, I'm not a billionaire.

Re: OpenAI departures: Why can’t former employees talk?

#379

I have some experience with rich people who think they can just put whatever they want in contracts and then stare at you until you sign it because you are physically dependent on eating food every day. Turns out they're right, they can put whatever they want in a contract. And again, they are correct that their wage slaves will 99.99% of the time sign whatever paper he pushes in front of them while saying "as a cond…

> stare at you until you sign it because you are physically dependent on eating food every day Even lowest level fast food workers can choose a different employer. An engineer working at OpenAI certainly has a lot of opportunities to choose from. Even when I only had three years in the industry, mid at best, I asked to change the contract I was presented with because non-compete was too restrictive — and they did it.…

I am typically not willing to bet I can get back under health insurance for my family within the next 0-4 weeks. And paying for COBRA on a family plan is basically like going from earning $X/mo to drawing $-X/mo.

Re: OpenAI departures: Why can’t former employees talk?

#380

The best approach to circumventing the nondisclosure agreement is for the affected employees to get together, write out everything they want to say about OpenAI, train an LLM on that text, and then release it. Based on these companies' arguments that copyrighted material is not actually reproduced by these models, and that any seemingly-infringing use is the responsibility of the user of the model rather than those w…

if I slaved away at openai for a year to get some equity, I don't think I would want to be the one to try this strategy
Post reply on HN