We’re getting sued for the homes we built that have unreachable floors and outdoor facing windows placed inscrutably on load bearing interior walls but I feel really good about our legal defense
Debbuging, for example, is harder than programming. And the difficulty is related to the 'amount of code' you are debbuging.
This means it is pretty easy to verify the Copilot's output when he just spits out a couple of lines at a time. But you will have a pretty bad time when you ask ChatGPT for a complete service. Things will not work out and the time you take to fix it will be greater than what you saved in the first place.
That sounds like a nightmare, not an improvement. Really, think about it. It's the same shit as "self driving" cars that need continuous human monitoring. You're taking an relatively engaging task and replacing it with a mind-numbing slog that humans are particularly bad at.
Not to mention the skill of being able to "check work" usually flows from deep experience of "doing work."
> It's way faster to check if a Sudoku puzzle is correct than it is to solve the Sudoku puzzle.
Yeah, and which of those tasks do humans choose to do? I don't see many "100 Solved Sudoku Puzzles To Check" books in the bookstore.
That does not map to instantiating a construction plan from scratch.
"Oh, yeah, that was a fuckup but the AI did it, not me! What, I was supposed to catch it? Well, I blame the AI!"
This seems oddly specific, has this already happened in real life?
The article specifically mentions agriculture, where machine vision has barely been deployed, and the company that is making the automatic pest / weed killer using lasers and MV is still in a "trial".
A case of: When AI is your hammer, everything starts to look like a nail.
Sure, but before half the world decided that "AGI" meant what I always called "ASI", the "G" meant "General", as in it's a general-use tool, so the hammer/nail analogy isn't anything like as apt as when "AI" meant "plays chess and nothing else".
Meanwhile image recognition is already being used to identify weeds to avoid spraying on the entire crop. Combines are somewhat self driving.
> We are thought leaders in AI. Like us and trust us!!!
> Our NFX portfolio company EvenUp has used our “AI inside” idea. EvenUp is an AI-driven platform that uses medical records and legal data to speed up the process of personal injury law.
> You should invest in EvenUp or become a customer. You should give us your money because we r smart and on trend!!!
We've had lots and lots of economic growth... but at what cost?
Youtubers who grow their own farm can only do it because it's their full time job. Or all their time is spent on the farm. Growing food is not easy, and would cost way more home grown.
Not something to romanticise. Nor would going out to the woods help, the carbon footprint would be so much higher than being able to have a expensive house with mega efficient everything.
Can't wait to never be able to contact a company again because they want me to talk to their ai shitbot that's designed to trick me into going away and not bothering them.
Don't use it.
>I want to build my own house, grow my own food, make my own shoes.
AI is not stopping you from doing these things.
Show me actual results today, not tomorrow's promises. At least TFA cites some results.
But I have a hard time extrapolating that out further. I suppose I'm not cut out for thought leadership. :)
I installed the new Ubuntu LTS and was trying to install a cargo package and was running into issue. Google/Bing search took me on a tangent. I put in the error in ChatGPT 3.5, got exact apt install to solve the issue and it worked when tried.
A million of these a day is going to have one heck of an effect on the planet.
I am not saying AI fixes the search but the Search Results are hit and miss any way. The real low hanging fruit are the domains, where the existing system is hit and miss.
The AI long case is, that People are also glorified LLMs/Pattern-matching machines. People who can make Computers sing will make AI sing (most likely). For most domains, AI elevates their 'base line' from the existing floor.
landline -> desktop -> mobile is a progression of technologies ?
Maybe "AI" might be able to transform technology, but if thats the demonstration of an example, I don't think the progression works how they think it does.
The implicit assumption seems to be the latter technologies in the list are strictly superior to the former, which is completely false. The article says this:
> Technological leapfrogging occurs when an industry or market (usually an outmoded industry or emerging market) skips a step along the technology transformation chain.
> Instead of learning to use a personal computer and then a mobile phone, you skip right to mobile. In many emerging markets, mobile is the dominant computing paradigm.
That kind of leapfrogging actually seems like a massive handicap. Mobile phones have severe limitations compared to PCs as devices for productive work.
Push button legal contracts, great, we have those already, they are called templates. The problem is not the contract, its the litigation that follows the contract.
> Generative AI is an instantaneous push-button solution. It generates a legal brief or construction plan from scratch. Interacting with software is like chatting to a friend. You don’t have to re-learn your whole process – you simply remove tedious tasks from your to-do list.
Ok ok ok, but thats a could. I mean I can ask a chatbot to make a contract for me, but is it legal, does it have a back door in it that allows something stupid to happen?
Same with Visas, the hard part isn't the form filling, its getting past the burdens of proof.
Where Legal chatbots might be useful, some day, is flagging for odd clauses.
_eventually_ I can see that "AI" will help automate a bunch of legal stuff, but its out of reach of current LLMs, as their inference of implications from text is sketchy at best. Moreover, new laws/contracting terms have less training data, which tends to bias the output in the wrong direction.
