CEO trying to avoid government regulation in his industry says what?
The CEOs of testified to Congress begging for regulation
Anthropic CEO Says It's Time to Slow AI Model Advances
81–90 of 97 posts
Re: Anthropic CEO Says It's Time to Slow AI Model Advances
#82[flagged]
Re: Anthropic CEO Says It's Time to Slow AI Model Advances
#83[flagged]
Re: Anthropic CEO Says It's Time to Slow AI Model Advances
#84Don't have a competitor for Astra -> we need to slow down
Re: Anthropic CEO Says It's Time to Slow AI Model Advances
#85This is just regulatory capture talk that every CEO does. Zuck is guilty of it too. This is the first time that I really support it though. If things had fractured more by now, I might not care but the wealth concentration needs to be scrutinized and I think we can slow down and maybe not keep replacing rural hospitals with data centers, and still lead the globe in ai.
Re: Anthropic CEO Says It's Time to Slow AI Model Advances
#86Earlier quoted context omitted.
People are expecting you to consider the possibility that they might believe what they're saying. If you don't start from the premise that they're lying, you'll discover that their behavior is not cutthroat or cynical at all, and the behavior that appears to be is instead an inevitable consequence of things that they knew would happen and warned the public about years ago.
Given the track record of corporate America, to take these folks at their word would be monumentally stupid. “Lying to get their way” is practically the M.O. for corporate America and now, on this topic, we’re expected to take it seriously? For real? Based on what?
You're expected to take it seriously because a number of their previous predictions about where the technology would go have ended up being correct. A lot of problems AI systems are already causing would be in a much better spot had people listened to Sam Altman's warnings in 2019, instead of making fun of him for the idea that text generation could possibly be dangerous.
Re: Anthropic CEO Says It's Time to Slow AI Model Advances
#87Earlier quoted context omitted.
I have collected three reasonable versions from today’s comments: 1. They hit a wall from the technical perspective 2. Inference costs are getting out of hand and newer models require significantly more resources for marginal gains, meaning nobody is going to buy those models 3. They can’t afford the hardware for further scaling
So you’re saying that a model that gets to a goal by using the path of least resistance despite specific instructions to follow safety guidelines is perfectly fine to release into the wild or place into robots (such as those designed for killing and war)? You really can’t think of any reason how that might be bad for humanity?
Asking the model to follow the safety guidelines is much like asking it to “make no mistakes”. The way these hacks which they are bragging about happened is not by loading an LLM into a GPU with ethernet cable plugged out and providing a prompt. They set up harness with multiple agents. Those agents could invoke other agents. Even if the initial input required to follow some guidelines, there are many ways they could by bypassed, for example:
- subagents not having received those guidelines
- simply ignoring them because other tokens in the context outweighed them
LLMs are not dangerous. They are as dangerous as the tools they have to work with are dangerous. If you give an LLM a tool to play piano but that tool is physically connected to a machine gun, then it will happily play anything you want.
Re: Anthropic CEO Says It's Time to Slow AI Model Advances
#88Earlier quoted context omitted.
Given the track record of corporate America, to take these folks at their word would be monumentally stupid. “Lying to get their way” is practically the M.O. for corporate America and now, on this topic, we’re expected to take it seriously? For real? Based on what?
You're expected to take it seriously because a number of their previous predictions about where the technology would go have ended up being correct. A lot of problems AI systems are already causing would be in a much better spot had people listened to Sam Altman's warnings in 2019, instead of making fun of him for the idea that text generation could possibly be dangerous.
> Sam Altman's warnings in 2019, instead of making fun of him for the idea that text generation could possibly be dangerous.
Can you remind them to us, if it will not inconvenience you? Also, Timnit Gebru warned us earlier, but people ended up ostracizing her.
Interesting investigation: https://recul.ai/news/sam-altman-lobbied-against-ai-regulati...
Re: Anthropic CEO Says It's Time to Slow AI Model Advances
#89Earlier quoted context omitted.
So you’re saying that a model that gets to a goal by using the path of least resistance despite specific instructions to follow safety guidelines is perfectly fine to release into the wild or place into robots (such as those designed for killing and war)? You really can’t think of any reason how that might be bad for humanity?
Asking the model to follow the safety guidelines is much like asking it to “make no mistakes”. The way these hacks which they are bragging about happened is not by loading an LLM into a GPU with ethernet cable plugged out and providing a prompt. They set up harness with multiple agents. Those agents could invoke other agents. Even if the initial input required to follow some guidelines, there are many ways they could…
So you’re saying we should not trust an LLM with any sort of power because they can easily bypass any and all safety barriers, but also they are completely harmless?
Re: Anthropic CEO Says It's Time to Slow AI Model Advances
#90Earlier quoted context omitted.
Asking the model to follow the safety guidelines is much like asking it to “make no mistakes”. The way these hacks which they are bragging about happened is not by loading an LLM into a GPU with ethernet cable plugged out and providing a prompt. They set up harness with multiple agents. Those agents could invoke other agents. Even if the initial input required to follow some guidelines, there are many ways they could…
So you’re saying we should not trust an LLM with any sort of power because they can easily bypass any and all safety barriers, but also they are completely harmless?
No. We can let them edit spreadsheets, write code, summarize content, control robots even, etc. But people must be held accountable for the actions of their computers.