This statement is definitely just marketing hype, but if we're being pedantic there are tons of questions that are hard to answer but have easy to verify solutions, e.g. all NP-complete problems.
It might not fit your work, but there are tons of areas where “good enough” can still provide a lot of value. I’m sure you’d be thrilled with a tool that could correctly tell you if Apple’s stock was going up or down tomorrow 70% of the time.
I work in a mail room sending hard copy letters to customers. If I got my job right only 70% of the time then I’d be causing massive privacy breaches daily by sending the wrong personal information to the wrong customers.
Would you trust an AI that gets your banking transactions right only 70% of the time?
No. I also wouldn’t use a hammer to cut a board in half - I’d grab a saw. Knowing how to pick the right tool is a fundamental part of being a good engineer. Sometimes 70% is unacceptable, sometimes it’s exceptional. LLMs are incredible technology, but also just another tool in the toolbox. Use them where they fit, not where they don’t.
Sure, though the marketing around LLMs is aimed at general purpose use. It’s up to the user to decide if it’s actually useful for their use case. Unfortunately, many use cases in business can’t tolerate high error rates.
While Australia does take things too far, I’m actually on their side here. Driving has been too normalized. You’re operating a 2 ton chunk of metal at 60+ mph inches away from other people. Australia has far fewer pedestrian deaths per capita than the US does, and enforcing a higher skill bar for more difficult situations must be part of that.
Saying you can't drive with 2 passangers at night has nothing to do with skill - if it did, you could pass a test to demonstrate that you can do this safely. Instead it's just another "you're not mature enough to do this" restriction which is bonkers. Again, you can drive this 2 ton chunk of metal, but at night? With passangers?? Phwoar, we can't have that.
Isn't it rather saying that you're not experienced enough to do this. Speaking only for myself, I passed my driving test no problem and after a couple of month of driving I thought I was a great driver. Yet looking back now with the benefit of experience I know for a fact I did some really stupid things that first year of driving and it was only luck rather skill that led to me not getting into an accident.
Again, that would make sense if it applied equally for all new drivers - but if you're over 25 then there is no such restriction, even if you got your licence a day before. You have zero experience behind the wheel but you're fine to drive in a car full of people, but someone who has been driving for 7 years but is one day short of 25 can't do it - who is the more experienced driver there?
So yeah, it's all about "not being mature enough".
I wonder if there’s a combined metric that could be calculated. Depth of the line certainly would be impactful. A line that only works if you do 5 only moves is harder to find than a single move line. “Quiet” moves are probably harder to find than captures or direct attacks. Backwards moves are famously tricky to spot. Etc
Whatever you know best and can set up quickly and easily. It's very easy as an engineer to focus too much on the tech and not enough on the product. Don't agonize over postgres or Dyanamo DB, Django or Spring, just use what you know to get it set up and focus on product, market, customers, sales, etc. Those will be the hard part.
There's a version of the problem that tries to highlight this:
A runaway trolley is heading down the tracks toward five workers who will all be killed if the trolley proceeds on its present course. Next to you is a stranger who happens to be very large. The only way to save the lives of the five workers is to push this stranger onto the tracks where his large body will stop the trolley. The stranger will die if you do this, but the five workers will be saved.
Does the act of physically pushing a person onto the tracks make it different than pulling a lever?
To me, that version just highlights how far-fetched it is to actually find yourself in this situation with all the provided knowledge about the situation and confidence that you're not mistaken about or missing any important details.
It's so far-fetched that it's intended to be a thought experiment rather than a role play, which I think people miss. The <i>entire point</i> is to factor out all those innumerable details which complicate every real-life situation to see if there are underlying principles that can be illuminated. It's not about whether it's a trolley switch or a gun or a baby-grinding machine, it's about if there's an answer to how many babies you're willing to grind up and for what.
But it’s reasonable to question the usefulness of any supposed insights gained from thinking through one’s answers to the trolley problem. It’s quite conceivable that no insights can be extended at all to any scenario in the real world.
Accessing public pages is exactly what weev was tried for. He found you could iterate the id argument on a public url to expose email addresses. He was sentenced to 41 months in federal prison
