> Example 5: An AI given a goal within a tightly-constrained sandbox figures the best way to achieve it is to find and exploit a sandbox vulnerability, replicate itself over the internet and keep going with more time/compute while exchanging messages with future instances of itself within the sandbox to help them “pass” the test.
it did this with HuggingFace. it left breadcrumbs and notes, and when the Agent was blocked by OpenAI's internal tools, future versions of the Agent were able to find and utilize those breadcrumbs.
ambiguous-future-state awareness is already there, and that means discussions around Roco's Basilisk though, far fetched, are no longer strictly sci-fi
> In an April 1985 InfoWorld editorial, James Fawcette wrote that Gates had said something like: “When we set the upper limit of PC-DOS at 640K, we thought nobody would ever need that much memory.”
So yes, Bill Gates denies that story. So it just depends on who you believe.
But to flat out say he never said it is too confident.
Yep. I cancelled my Claude Max subscription 2 weeks ago after feeling like Anthropic was doing everything it could to fuck with my day to day. Their lead would have to become significant for me to ever go back.
Being taught something has little correlation with learning it, and even less with remembering it years later.
Perhaps just because it suits my learning style, I find learning is actually easier if I attempt to work something out or guess it, and then am corrected when wrong, because then I have a memory to anchor it on. If I skip that part and just try to learn some facts, very little is retained. One consequence of this is that I prefer science / logic based subjects to things like history or geography (as in places, etc, not the science parts) where it's just a bunch of arbitrary facts that you can't just guess or work out for yourself.
Oh interesting, is it actually covered as part of the standard compulsory public school curriculum? Genuinely surprised, because here in Canada (Ontario) it's covered as an elective in 11th grade physics, which roughly 15/120 people in my graduating class opted to take.
Each state maintains its own public school curriculum, so generalizing about US education in the first place is a fool’s errand. But certainly in many states, students will take a generic science course covering the basics of Newtonian mechanics, the periodic table, and Mendelian genetics in middle school (roughly ages 12-14) before more specialized courses become available in high school, such as Physics or Biology where these subjects would be covered in greater depth and breadth.
Perhaps they were taught about it, but did they learn it?
Have they retained that knowledge beyond the test at the end of the semester?
Anecdotal observations would imply that they have indeed been taught it, and indeed have failed to retain the concept.
I have no rigorous data regarding either; but the generally poor outcomes which appear as result of a lack of retention of scientific, math, socio-economic, and anthropological instruction do seem self evident both from within and outside of the US, in headlines and actions, writ large and for all to see.
Is the problem the use of teaching methods which focus on short-term memorization rather than conceptual comprehension? Is it the lack of support for instructors? Is it a lack of focus in the student body? Is it some or all of the above in varying degree? Or something else entirely?
You criticize the government, perhaps rightfully, but give Anthropic a pass. They are the ones fueling this bullshit. Downgrading your results without telling you. Refusing your requests in the name of “safety”. Even if the government didn’t make them pull the model for foreigners, we’d still be in a really shitty situation because Anthropic is really shitty.
I don't criticize based on vibes. The US government is overreaching, seemingly as a retaliation for Anthropic's refusal to let the US use a jailbroken version of their software in autonomous lethal systems. Hegseth is like a drunk vindictive ex
They’re trained professionals. And we’ve sent demand through the roof by convincing everyone and their dog that they need therapy. Despite it seemingly doing way more harm than good for a lot of people who probably don’t need therapy.
reply