Yeah, this is what I was looking for. In the past, Google searches have showed up with nothing. There wasn't a good way to hire someone for a small amount of work. Bookmarked!
Everybody talks like they have a lawyer just sitting outside on their patio and all they have to do is give them a couple hundred bucks to look over a document here and there. How does one find a lawyer that will work piecemeal like that?
Right. Those are tied via evolution to the dynamics of the physical world. We can simulate the physical world and learn from that, but there needs to be a there there. Language assumes the listener already has that understanding.
Language is a representation medium for the world, it isn't the world itself. When we talk, we only say what can't be inferred because we assume the listener has a basic understanding of the dynamics of the world (e.g., if I push a table the things on it will also move). Having an AI watch youtube and enabling it to act out what it sees in simulation would give it that grounding. We are heading that direction. So, I agree ChatGPT is awesome. I don't believe it understands what it is saying, but it can if it trains by acting out what it sees on Youtube.
I think we will see AGI. But for the AI to be robust, it has to interact with the world, even if it is a simulated one. We need to build an AI that knows what a toddler knows before we can build one that understands wikipedia.
The reinforcement learning was on giving the right answer, not on interacting with the world. But there is movement in the right direction with https://ai.googleblog.com/2022/12/rt-1-robotics-transformer-...
and other RL stuff. (RT-1 isn't RL but there is other related stuff that is)
It's because ChatGPT's knowledge doesn't come through interaction with the world. Words mean things to us because they point to world interactions. I saw a demo where ChatGPT could create a VM. It could be trained to interact directly with a VM, send commands to an interpreter. In this case, it would understand the response of the VM, although it wouldn't understand the design behind the VM, because humans did that based on interaction with the physical world.
Wouldn't that just continue to expand a bad practice? If drivers got tips, Amazon would lower their wages, and we'd just have another domain where we have to make extra decisions.
This all sounds good, but to me it highlights the downside of not enforcing laws. If you force people to hire private security, it's going to be hard to ensure that they are all nice and reasonable. Also, it's just a bad way to live with all of these private security forces everywhere.
It starts out strong, identifying bullshitters on the left and right, but if Obama and Pinker are bullshitters, I'm not sure the label has much meaning. Which is a shame, because I too have been surprised at how little the truth seems to mean to people.
I don't know, I've had pretty bad service recently at places where you tip. I still top 20% out of guilt, but it might be that service in general is poorer now.
It's hot, but 115-degree summers is an exaggeration. The high is usually around 100 degrees (still too hot, of course). I do much prefer the weather in California.
Speaking of all this, on my windows 11 machine I get McAfee popups, and for the life of me, I can't track down where they come from or how to turn them off. My son has a windows 10 machine and has the same problem. Why does Microsoft put up with this? I only use windows when I have to because of crap like this.
One thing I like about Bank of America is their app and website are both really good. I have another account with a small bank and I continually have problems with the website being janky.
I keep hearing about how in the US enrollment is going to drop in a few years when the kids (not) born during the great recession of 2008-2010 hit college age. Interesting to see how that will play out with this administrative bloat.
I love the "Show HN: $X created by ChatGPT" We need people seeing what useful things can be done with this new hammer, things that haven't occurred to us yet.
It can generate a bunch of stuff automatically. What we need is the ability to test that stuff automatically, so we can close the generate-and-test loop. If we can test automatically, we can then have this thing hillclimb to usefulness. I mean "test" in the sense of "evaluate whether the output is good."