[t0mas88]

> OpenAI, meanwhile, has been attempting to quell the backlash against its deal with the U.S. government, putting out a blog post claiming that “our tools will not be used to conduct domestic surveillance of U.S. persons,”

As a non-US person, that sounds far more concerning than no statement at all. Because if their tools weren't used for surveillance against Europeans they would have said so as a marketing message...

[adrianN]

With n-eyes agreements it’s quite meaningless anyway. Whatever passport you have, somebody spies on you and sells the information to your government.

[shimman]

It's also meaningless because we know governments get around these "agreements" by buying data from third party companies that bought the data from OpenAI. The only way to stop this is to legislate it out of existence.

[drob518]

Yep. “You spy on mine, and I’ll spy on yours, and then we’ll share info.”

[kakacik]

I wouldn't give them any free pass and just give up, its highly amoral and inhuman behavior. Modern form of racism but based on passport.

You have this one? You are subhuman, treated as such and you have very limited rights on our soil, we can do nasty things to you without any court, defense, or hope for fairness. You have that one? Please welcome back.

Sociopathic behavior. Then don't wonder why most of the world is again starting to hate US with passion. I don't mean countries where you already killed hundreds of thousands of civilians, I mean whole world. There isn't a single country out there currently even OK with US, thats more than 95% of the mankind. Why the fuck do you guys allow this? Its not even current gov, rather long term US tradition going back at least till 9/11.

[lavezzi]

> "We have these two red lines... Not allowing Anthropic's AI to perform mass surveillance of Americans, and prohibiting its AI from powering fully-autonomous weapons..."

Anthropic literally said the same, but seem to be getting positive PR.

https://www.cbsnews.com/news/ai-executive-dario-amodei-on-th...

[gruez]

It's not "literally the same".

https://www.lesswrong.com/posts/FSGfzDLFdFtRDADF4/openai-s-s...

[fluidcruft]

The difference is that Anthropic actually dotted the i's and crossed the t's whereas OpenAI fell for the weaselwords and is now desperately trying to renegotiate.

[beachy]

OpenAI didn't fall for anything, they knew exactly what they were signing and went ahead anyway, then started gaslighting people about what they had signed.

For a lot of people (me included) the lack of integrity and the gaslighting is what has soured them on OpenAI, rather than them signing up to build surveillance and weaponry.

To non-US citizens, all AI companies are as dangerous as each other, OpenAI just really botched the optics here.

[spwa4]

It's amazing how bad FANG executives are at even knowing what a normal moral thought would be for average people ...

Plus, you know, you'd think they'd ask their cleaner or baker or something. Or hire someone.

[clcaev]

Executives are certainly capable of understanding moral/ethical concerns.

Around 2005, a Yale Psychology PhD candidate asked me to write a web-based survey instrument with various questions, some on complex but straight forward business questions (the controls) and others with moral/ethical aspects. Senior executives participated and they answered similarly to rank & file, often completing the entire survey much faster. What they didn't know -- we were tracking how long they spent on each question. Questions with moral/ethical concerns took senior executives relatively longer than the rank & file.

Late Addendum: Sorry that I don't recall the author/paper. The survey population spanned multiple industries representing many Fortune 500s, including huge tech companies. The survey was the same for everyone. The questions were story problems from business and law school case reports. The participating companies were anonymized on our end. We provided HR departments with survey link; only subject rank (not identity) was collected. Survey was voluntary, with informed consent according to IRB approval.

[mjburgess]

You would also need to control for the degree to which people had a stake in the outcome (ie., virtue signalling).

Since executives have to make decisions where choosing the moral option may impose an economic (or operational) cost, this requires thinking through the actual choice.

Morality for the "rank and file" is just a signalling issue: there's nothing to think through, the answer they are "supposed to choose" is the one they do so, at no cost to them.

[clcaev]

I hope the addendum helps clarify.

This study showed executives spent relatively more time on questions with moral/ethical concerns. Perhaps the control questions were more similar daily work and hence familiar, while there were fewer encounters with questions having moral/ethical concerns. Perhaps executives decided more care was required for these questions to ensure people were not hurt.

Getting back to the grandparent post, executives are certainly aware of situations with moral/ethical concerns and need not consult their barber to answer them.

[Timon3]

"Rank and file" employees choosing to prioritize morality very, very frequently pay real costs for doing so - with a much larger personal impact than executives feel.

[mjburgess]

Only in very rare circumstances where the obvious answer and their procedural work dont align.

When making an operational decision that affects the direction of the business, morality is almost always a concern -- even at the level of "do our customers benefit from this vs., do we?" etc.

[Timon3]

Where do you get the idea that those circumstances are "very rare"? Workers are being asked to break rules and do unethical things all the time, and you're pretty much guaranteed to pay a personal cost if you refuse.

Meanwhile morality is almost always one of least important factors when making operational decisions.