junofan

31 ms·

511.
by junofan 3y ago
You have evidence Sam lied, like their statement implies? How do you know the action was in furtherance of their charter instead of over petty grievances?
512.
by junofan 3y ago
I don’t see how that factors in. What matters is OpenAI’s enterprise customers reading about a boardroom coup in the WSJ. Completely avoidable destruction of value.
513.
by junofan 3y ago
Why would any employee put up with that? Why not go work somewhere else that better aligns with what you want?
514.
by junofan 3y ago
They are currently in talks to rehire him. It would seem they wish to rehire him?
515.
by junofan 3y ago
Their actions were reckless and irresponsible. They are currently picking their successors. The incompetence is staggering regardless of motive.
516.
by junofan 3y ago
I think you’re misreading the situation. Nobody thinks Sam’s a god. A lot of people now rely on OpenAI for parts of their livelihoods and would like a steady hand. The board’s actions don’t seem fit to govern an HOA, much less an important
517.
by junofan 3y ago
I just don’t see what value this brings to the table. Sounds like Meta might not either. We’ll just have to leave it at that.
518.
by junofan 3y ago
Come on, that was fun and engaging. A floating rock is pretty harmless.
519.
by junofan 3y ago
I don’t think I misread anything. I wasn’t talking about the synthesis steps. I don’t see any additional risk here. All the information presented is already widely available AFAIK. The handwringing damages credibility.
520.
by junofan 3y ago
Assuming a historical figure wasn’t racist feels like a bad default. I don’t think it’s necessary to point out racism in every hagiography.
521.
by junofan 3y ago
I just don’t see the danger. There isn’t anything you couldn’t find on 4chan in a few clicks. And the bioweapons example is a pointer to RefSeq? Come on. These efforts just don’t stand up to scrutiny. They risk appearing unserious to peopl
522.
by junofan 3y ago
I think you might want to take a look at some of the history here, and particularly the cyclical nature of the AI field for the past 50–60 years. It’s helpful to put what everyone’s saying in context.
523.
by junofan 3y ago
Nobody’s taking those instructions and building out a lab successfully who doesn’t already know what they’re doing haha. What have the safety people stopped so far? That’s where I’m struggling to see the point.
524.
by junofan 3y ago
No? That’s been the goal of NLP and information retrieval research for decades.
525.
by junofan 3y ago
There’s no way to build out a wet lab with a chatbot for weapons production. It’s just not feasible. Your example is a fantasy. Fraud is covered by the legal system. I don’t know anything about nerve agents.
526.
by junofan 3y ago
Sounds more like a religion than a well-defined business objective.
527.
by junofan 3y ago
That thread is simply unhinged. There is no terrorist with a wet lab who outright refuses to read papers and instead relies on a chatbot to work with dangerous agents.
528.
by junofan 3y ago
I’m not trivializing risks. I’m characterizing output. These systems aren’t theoretical anymore. They’re used by hundreds of millions of people daily in one form or another. What are these teams accomplishing? Give me a concrete example of
529.
by junofan 3y ago
Not seeing the sexism. I think the AI safety community is a little early and notoriety therein probably isn’t sufficient to qualify somebody to direct an $80 bn company.
530.
by junofan 3y ago
I don’t think everyone appreciated suddenly taking part in an iterative scientific process without consent.
531.
by junofan 3y ago
We’re talking about making sure chatbots don’t say “nigger,” not bioweapons. At some point you need to trust that the people using the tools are adults.
532.
by junofan 3y ago
It doesn’t sound like Sutskever is running anything. OpenAI reportedly put out a memo saying they’re trying to get Sam and Greg back: https://www.theinformation.com/articles/openai-optimistic-it...
533.
by junofan 3y ago
The board folded. > Do you think that the term sheet holds OpenAI liable for changes in Microsoft's stock price? There’s nothing binding on a term sheet.
534.
by junofan 3y ago
It’s over. He’s back.
535.
by junofan 3y ago
It doesn’t sound like either of us have all the facts. We’ll leave it at that.
536.
by junofan 3y ago
That’s unfair. The issue is poor governance. Why would anybody outside OpenAI care how much money they make? The fact is a lot of people now rely in one way or another on OpenAI’s services. Arbitrary and capricious decisions affect them.
537.
by junofan 3y ago
Various newswire-type things on Twitter reporting Satya will do what it takes to make sure Sam’s at work on Monday.
538.
by junofan 3y ago
There’s no owners. No ownership interest to sell. The board answers to the courts.
539.
by junofan 3y ago
I don’t follow. You’re saying PPUs granted by the for-profit don’t carry a reasonably expected benefit? And that a parent can’t tortuously interfere with a subsidiary? Both are factually untrue. I don’t see the wiggle room here that you do.
540.
by junofan 3y ago
Really? I don’t know of a positional encoding scheme that’ll handle this.

More ›