> yes! and we need to add the death penalty to the list of punishments for children, because what matters is not the size of the crime, but the harshness of enforcement, that's the real deterrent, arewerite!?
Child death penalties probably aren't something that you actually want, right? You NEED them. Nothing gets people to take a step back like capital punishment for misdemeanor crimes. And with great power comes great responsibility.
However, I get where you're coming from. A thought: Speeding is a pretty tiny crime but it needs to be punishable by the death penalty on the first offense. We'd never have to worry about that criminal driving dangerously again.
It actually stones 2 birds with 1 kill. That speeder might actually be a serial speeder. Serial offenders sometimes escalate their crimes over time and they never, ever stop until they're not just dead, but also decapitated. The serial speeder that drove 5 over yesterday might drive 7 over today. Tomorrow may even get all the way to 8 over the limit. Eventually they'll get to ramming speed. After a serial speeder gets to ramming speed, they will never accept a slower speed; they will ram other cars and die first. This would have been prevented by simply taking advantage of the power of 1st offense: death penalty.
For both time efficiency purposes and the perception it would bring, judges should be walking around with 2 fully automatic assault rifles tucked under their robes. They already line up before the verdict is read, then dump both mags as soon as the jury says "guilty."
There's an overly lax legal system and no signs of a "Death Row Children's Fun Zone". The reality of that approach is that kids have the freedom ( and enough tokens) for basic white collar crimes. If enough of that happens, money will get canceled forever. Then by the time next Tuesday comes around we'll wake up on an Earth with everything regressed all the way back to a pure barter economy. This is why we can't have nice things.
> I also disagree with the essentialist position that the best education possible must be pure-human provided. Maybe if cost is no object and you have a 24/7 human tutor on speed dial for each kid, that would be superior to any AI-assisted form. But it seems pretty obvious to me that human+ai could deliver better results than human alone, in the same way that AI is very clearly enabling GPs to broaden their diagnostic ability, spend more time connecting with patients, and reduce their paperwork toil, when deployed with care.
You seem to think you're making some kind of valid argument here by backing up your opinion about one thing by providing evidence of something unrelated aside from both things happening to involve AI. I'm guessing only reason it seems obvious to you is because it's not obvious to you that what you're disagreeing is something different than what's going on to begin with.
I haven't seen anyone say the best education would be pure-human provided. The entire point of the article is that the effect AI has had on the education of children so far has not been beneficial.
The fact that AI useful for GPs for getting relatively simple yet time consuming administrative work taken care of has nothing to do with children's education. GPs aren't doing paperwork to develop their fundamental reading and math skills. Less time dealing with the non-medical portions of running a practice means they can spend more time with their patients. Spending more time with their patients allows them to better understand their symptoms and make a better informed, more accurate diagnosis even if AI is never used in a patient-facing capacity. In no way does that indicate that AI is currently a suitable to deploy in education.
I don't know if you used AI to learn about what you were arguing with and gather information to back up your argument, but if you did then that's another example of why there's so much concern about little kids using it. Eventually it will be worked into something without such a double edged nature when applied to early learning. Everyone already knows it can be an educational boon for people who have the skills to use it properly. So far, statistics indicate students do not have those skills. The rate of failure among undergraduate students is only continuing to accelerate.
I remember that whole "expect 30 to 60 min a night" nonsense. I always assumed that was a scare tactic or something. I mean, in high school i was awake until around 4 am playing games every night so I'd wind up using half of each school day compensating with well-timed napping. Even then, the only homework i had to actually take home was the stuff meant to be turned in a few weeks later like tests and projects. Sure, we had daily "homework" assigned; there was also more than enough time during lectures to finish it. The only students that i really ever saw complain about having tons of homework were the ones that didn't put any effort into learning stuff at school to begin with.
> The net outcome there is going to be highly negative.
Agreed. With the way LLMs often are quite artuculate and confident sounding, it's only a matter of time until kids will be suspicious of actual people that are articulate and confident from putting effort into improving their speech and speaking about things they are right to be confident about due to being proficient in the subject.
Either that or kids will see that LLMs aren't "getting punished" for being "convincing liars" and will grow up convincing the socially correct way to interact with others is by speaking to them respectfully while going all in on long-game gaslighting.
> belch out thick black lib-ownin' ball-crushin' smoke, damaging their own engine in the process.
The black smoke belching is caused by a fuel rich air mixture that leads to incomplete combustion and is pumped out of the engine as a burning cloud of smog. While this can cause problems with carbon build-up in the combustion chambers, it's not especially difficult to clean that out before the engine suffers permanent damage. Most often the worst case is fouled spark plugs, which are wear items meant to be replaced at regular intervals.
The real damage occurs in the exhaust plumbing, typically rendering the oxygen sensors inoperative and cooking the catalytic converter. However, a common way to induce black smoke is by hardwiring the oxygen sensors to report the engine needs more fuel, so it's not like the oxygen sensors were typically working anyway.
Likewise, in many parts of the country where this sort of "coal rolling" is common, it's also common for the catalytic converter to be forcibly removed from the vehicle and pawned off for the platinum mesh inside by a couple meth heads looking for their next fix, which has the side effect of preventing it from being damaged. The missing portion of the exhaust pipe is frequently repaired by welding in a straight pipe, allowing the black smoke to flow more freely while also making the exhaust notably louder and adding the scent of immediate global warming (note: this is pretty much super illegal in most states).
NASCAR has a tendency to fight changes to the drivetrain technology, so a hybrid might take a while still. The regulations banned fuel injection up until 2012, required transmissions to be 4 speed manuals until 2021 and will probably keep the 2 valve pushrod valvetrain and 90° V8 layout til the end of time. If they ever get around to implementing hybrid powertrains (they've been grumbling about the issue since 2024), they aren't going to be there to save gas. Racing hybrids are designed primarily to recover kinetic energy in deceleration, store it for short term and use it to boost acceleration. There are designs for race hybrids that don' t even use batteries or utilize electrical energy for the hybrid part; instead storing it mechanically in flywheels to avoid the inefficiency of converting from mechanical to chemical energy and back.
> What's destabilizing the industry right now isn't vulnerabilities AI introduces into new code; it's a flood of sev:hi vulnerabilities in existing code, not introduced by AI but discovered by it.
Vulnerability discovery has essentially moved to a "proof of work" computation model with AI that has some similarities to crypto like BTC or ethereum 1.0. I don't see any reason a well funded adversary couldn't use this same process on open-source code to develop exploits. I'm sure AI would be happy to try and create exploits from the results rather than fixes.
This sort of proof of work has a notable difference from crypto in the asymmetric nature of what each side is targeting. In crypto, each miner was attempting to find a solution to the same problem and they would all move on to a new one once a solution is found. However with AI vulnerability scanning, the non-deterministic nature means an adversary is likely to find different vulnerabilities. Even if it doesn't, the adversaries have a different post-discovery workflow (i.e. probably less compute intensive aka cheaper due to only needing one viable exploit to win) than the software maintainers do.
Considering it's possible both the adversary and their target could both do all this while running Claude puts Anthropic in a real "Merchant of Death" position.
> I understood your comment perfectly fine. I'm asking which graduates of which colleges you were referring to.
They are referring to MOST graduates of MOST colleges. This is a deliberate overgeneralization about the nature of post-secondary education meant to highlight how it's frequently viewed solely in terms of completion rather than with regards to any skills or knowledge gained from it.
> 4GB, $0.10 (whatever the HD price) that is the equivalent of a High School level intelligent brain that can perform many cognitive tasks for free?
This is better than my current solution of an actual human with masters degreed intelligence performing all my cognitive tasks for free how? I mean, i'm the first to admit i'm extremely lazy and even i'm over here like "really??"
> I don't think so. I know for a fact that search terms are a minefield of gotchas and hacks caused by product decisions that reflect ad-hoc negotiations with partners and sellers. It's an unstable equilibrium of partners trying to shift attention to their products in a certain way. I think that calling this fragile equilibrium optimized has no bearing with reality.
You think a crude, unoptimised "minefield" is the route that leads to something as delicate as a "fragile equilibrium?" I don't see something as carefully balanced as your unstable equilibrium even being something that could exist without the processes involved having been refined down to a science. The only real alternative that meets your narrative would be that this is an industry that runs entirely on hope and luck (and enough human sacrifices to keep ample supplies of both on hand).
Not sure you're aware but you initially sound like you disagree with the post you replied to, only to follow up by enthusiastically reiterating that author's words as if in agreement.
You realize what shoppers and vendors each consider to be "good" e-commerce sites are fundamentally opposed concepts?
"if a value-aligned, safety-conscious project comes close to building AGI before we do, we commit to stop competing with and start assisting this project"
The tried and true technique of speeding up completion of an established project by adding more people; well known for the tendency to fail successfully.
The reason AGI couldn't do these is the lack of a suitable interface to the physical world. It would take a trivial amount of effort for these to be designed and built by the AGI. Humans could be cut from the loop after an initial production run made up of just the subset of these physical interface devices needed to build more advanced ones.
> utilizing a “hack it till it works”(tm) methodology.
Your post describes my coding perfectly. I don't have CS training of any type, never been formally involved in software development (recently started dabbling in OSS) and never used an LLM/agent for help (do use a local SLM for autocomplete and suggestions only).
Yet I can "code." I suspect a (pre-2023ish) software developer would likely tell me "go learn to code" if i asked for review. I don't know the formal syntax people expect to see and it has organization more typical of raging dumpster fires. Doesn't mean it's not code.
> If the light switch you bought, has a little daylight sensor on it, and turns off when the sun is out, and that's what it does.. you may not like that light switch. You might want one that "does what you want, because you paid for it!" but then you should have purchased a different one, or made a light switch you actually liked.
Not sure this analogy works as it gives prospective light switch buyers a choice of different light switch types. What google is doing seems more like forcing EVERY light switch to have daylight sensors, thus forcing you to save power (even if you're pro-global warming and just trying to do your part for the cause), then telling people with vision problems relating to suboptimal indoor illumination or suffer from sunlight frequency melting disorder or think they've got some other random "daylight makes life suck" bullshit to create a student/hobbyist account.
Ditto. Coding isn't what i specifically do, but it's something i will choose to do when it's the most efficient solution to a problem. I have no problem describing what i need a program to do and how it should do so in a way that could be understandable even to a small child or clever golden retriever, but i'm not so great at the part where you pull out that syntactic sugar and get to turning people words into computer words. LLMs tend to do a pretty good job at translating languages regardless of whether i'm talking to a person or using a code editor, but i don't want them deciding what i wanted to say for me.
Child death penalties probably aren't something that you actually want, right? You NEED them. Nothing gets people to take a step back like capital punishment for misdemeanor crimes. And with great power comes great responsibility.
However, I get where you're coming from. A thought: Speeding is a pretty tiny crime but it needs to be punishable by the death penalty on the first offense. We'd never have to worry about that criminal driving dangerously again. It actually stones 2 birds with 1 kill. That speeder might actually be a serial speeder. Serial offenders sometimes escalate their crimes over time and they never, ever stop until they're not just dead, but also decapitated. The serial speeder that drove 5 over yesterday might drive 7 over today. Tomorrow may even get all the way to 8 over the limit. Eventually they'll get to ramming speed. After a serial speeder gets to ramming speed, they will never accept a slower speed; they will ram other cars and die first. This would have been prevented by simply taking advantage of the power of 1st offense: death penalty.
For both time efficiency purposes and the perception it would bring, judges should be walking around with 2 fully automatic assault rifles tucked under their robes. They already line up before the verdict is read, then dump both mags as soon as the jury says "guilty."
There's an overly lax legal system and no signs of a "Death Row Children's Fun Zone". The reality of that approach is that kids have the freedom ( and enough tokens) for basic white collar crimes. If enough of that happens, money will get canceled forever. Then by the time next Tuesday comes around we'll wake up on an Earth with everything regressed all the way back to a pure barter economy. This is why we can't have nice things.