I’m an anarchocommunist, all states are evil.
Your local herpetology guy.
Feel free to AMA about picking a pet/reptiles in general, I have a lot of recommendations for that!
disagree, if they reimplement this as foss software and it crushes adobe it will be nothing but a good thing, sure ai overall is bad but ai was happening anyway and minimizing the harm of it by using it for actually good things for the foss movement is the least problematic possible thing to do with a highly problematic tech.
Not using this doesn’t help with that though? It is already released, telling them not to make it would help but it’s too late now?
the initial article claims ““AI” – chatbots that wake up, “set their own goals,” and “spontaneously” start hacking servers – is fake.”
an article saying that is your goalpost, prove me wrong, quote something.
Powerful embodied llm’s pose legitimate actual threats and you’d have to be delusional not to see it.
“I don’t presume Anthropic is designing anything to be ethical.”
Then why do they put any safeguards of any sort in? They put a lot of work into this, also this is openai.
“First, the chatbot is unpredictable by design. Randomness is built in.”
Randomness is built in but this behavior was not designed, it was unpredictable and random, which is, yes, unpredictable. The goal isn’t to make it unpredictable and temperature controls exist for a reason.
“Second, if they didn’t intend for bad things to happen, why didn’t an employee babysit it?”
They did they just weren’t paying enough attention, this was a benchmark. You have to have someone check the results to be useful.
I reject that rogue is a goalpost shift or inaccurate. I think calling this rogue behavior is accurate.
rogue /rōg/ noun
An unprincipled, deceitful, and unreliable person; a scoundrel or rascal. One who is playfully mischievous; a scamp.
It did act that way, no? I did not shift goalposts. Remember the quote was
They did not intend for it to hack those websites, it realized that was a way of achieving its goals even though it’s specifically designed to be ethical. That can easily be called going rogue and I see no issue with it. They did not design it to pass the benchmark by hacking the website the benchmark was on. They hardly really design llm’s.
As the article said you can replace rogue agents with unpredictable computer programs if you really want, I don’t see the necessity.
Regardless that’s a major goalpost shift.
“what it spits out is never really anything new”
this is just false
https://www.youtube.com/watch?v=Z7dH_Cxx75g
https://github.com/openai/ten-proofs
unless you would agree that humans never spit out anything new, as well.
the first article doesn’t say anything not factual, the second article has a direct response to that
“By now, if you’re an A.I. skeptic, you’re probably silently yelling at me for anthropomorphizing these systems. Go ahead, but feel free to replace “rogue agents” with “unpredictable computer programs””
the third article also says nothing like that.
the initial article claims ““AI” – chatbots that wake up, “set their own goals,” and “spontaneously” start hacking servers – is fake.”
an article saying that is your goalpost, none of these articles say that. Even the headlines don’t say it and that’s where they normally put the crazy claims.
ironically the claim that people claim that the AI chatbots woke up set their own goals and spontaneously started hacking servers seems to be fake.
Nobody is claiming any of the things this article is fighting against, the chatbot in the hugging face incident followed the instructions given to it and there were unintended negative consequences, anybody serious about this is trying to avoid worse unintended negative consequences not whatever the hell the article is talking about.
Give an example of a problem a human can solve that an llm could not, then.
It can perform necessary tasks that actually drive intelligence and isn’t derivative. See: those proofs.
You’re essentially arguing that it’s useless and just moves data around, this isn’t true.
You’re a little too caught up on that specific wording.
It is derivative and cannot perform any necessary tasks that actually drive intelligence
done.
also, the closed loop point is just false read my article?
Superintelligence may be more about pushing efficiency rather than what you’re saying, i’m sure they can rapidly iterate on making more efficient models.
Yes you can obviously use a closed loop system and scale up? it’s even becoming the standard? https://datacentremagazine.com/news/how-closed-loop-cooling-is-reshaping-data-centre-design
That is a massive goalpost shift. You went from can’t do anything useful to couldn’t have discovered the problems in one step?
Why does it even matter if it can’t discover the problems if it can solve them, exactly? And why could it solve the problems and not discover new ones?
Point 2 isn’t really true, closed loop systems exist and are becoming the standard.
Point 3 is just obviously false? https://techjournal.org/openai-astra-solves-open-math-problems
Have you considered who’s currently in charge of federal safety standards?
Pinta?
The following filter options can’t be translated into DNR rules:
strict1p, strict3p: whether a network request is same-origin as its initiator
Entity-based values for domain= filter option (webextensions #394)
redirect-rule=: the DNR API does not support redirect-if-blocked concept (webextensions #493)
Click-to-load embedded contexts through redirect=click2load.html filter option
Regex-based removeparam= modifier filter options
Exceptions for all modifier filter options are not possible
Workarounds exist for removeparam= and csp= by carefully tuning the priority property.
Many very useful regex-based filters used in uBO are not allowed, or are rejected by the DNR API (webextensions #344)
replace=, can't modify the response body (full support is only possible with Firefox MV2)
ipaddress=, can't use IP address as a condition to block/unblock (full support is only possible with Firefox MV2)
urlskip=, this requires programmatic steps to extract a URL from another URL (partial support, for strict-blocked pages only)
aside from that it is somewhat better than default ublock but that’s a pretty big list of issues.
You should really use endeavoros for this, manjaros team is wildly incompetent and does way too much (which is a problem since they can’t get the basics right)
The point is a complete and utter removal of social mobility
They were going to use it anyway. The notion that your boycott will accomplish anything when the tool is this unbelievably useful is insane, fight this at the regulatory level, furthermore the ecological impact is quite minor if you use Chinese ai due to the solar farms.