In construction those are all called tractors (as are Bobcats, cranes, pavers, earthmovers etc.)
Probably a semantic error, but you can find the backhoe on the wikipedia page for Tractor.
Yes, and we are way off topic. My bad there.
The broader point was that the level of AI to operate heavy equipment is far advanced from that of delivery driving and taxi driving.
There is automation in the industry but as yet nothing that I would classify as AI.
You may have seen it on the farm as that was begun years before this AI craze.
The combine drives itself around the field. But the combine doesn’t train itself and makes no decisions.
I mean, even as I was replying, I was thinking about how construction is a long way off. However as I continued to think about it, I think there’s a difference between taking a task over for a human and designing a task for AI.
We could be a century away from robots working on construction sites as they are today. (I picked that number out of my hind end.) However a construction site could be designed for AI. Look at 3D house printers.
Imagine you could hire humans to level the ground, and the AI could take over and do all the building its own way. You have to automate cars to work within the confines of crazy humans driving. In a construction site, just like a warehouse, you do not if you design it around the robots.
Elitist how? Saying that jobs that are invented are not real? To say that programmers, designers, writers, etc., do more real work than people who just warm seats? I don’t give any examples of the jobs that are pointless…
Because saying that some jobs are not ‘real’ jobs is quite insulting for those doing those jobs. Sure there are absolutely some people out there who a blagging their way through existence, doesn’t mean their job isn’t important though.
You are claiming there are hundreds of millions of jobs that don’t require human intelligence, creativity, or wisdom out there for LLMs to replace, I don’t see how that’s not insulting to those jobs, but I was not very serious about that comment, but rather pointing out how companies are trying to replace people who do jobs that LLMs can’t replace
If we’re talking about those same hundreds of millions of jobs that can be replaced by a glorified word guessing machine then that’s the insulting part I’d say
Also I think people confuse that tasks can be automated with that jobs can be automated, they’re not the same thing. You can automate writing an email, but you can’t automate the broader process of the job (or you can in various cases, but those jobs have already been)
What else would this mean? How is a job something an LLM can replace and not a job that doesn’t require human intelligence, creativity, and wisdom?
What kind of jobs do you imagine it can replace?
Read my comment again please, I explained that I wasn’t serious that there are fake jobs, only that LLMs can only replace made up jobs that don’t require human traits that LLMs don’t have
For example, if you said my job could be replaced by an LLM I’d be deeply insulted, and if you think they can replace people in the vast majority of jobs you don’t understand those jobs IMO
Just because a job can be replaced with AI doesn’t mean it should be. By that logic you could replace all Lawyers with AI. All the AI needs to do is understand the letter of the law and its interpretation in the end. No intelligence is required, no creativity is required, and no wisdom is required, if we follow YOUR logic.
I think that there are quite a few lawyers that might have something to say about that though!
You took my statement to mean that AI should be replacing jobs that are not ‘mentally’‘ taxing. The statement was in no way written to mean that, quite the opposite in fact. It was to highlight how many jobs could potentially be at risk.
So can I have an example of a ‘made up’ job then please? Because in my world view there is no such thing and its quite insulting to say otherwise.
Is a taxi driver a ‘made up’ job? There are plenty of examples of trials of automated taxis currently in use around the world. Are all taxi drivers lacking human intelligence, creativity, and wisdom?
I think maybe you should stop further digging the crater size hole you have made for yourself.
You are not reading at all what I’m saying if you think that’s my logic… So I’ll just give up trying to respond to you, either that or you have no idea what a lawyer does if you think it doesn’t take creativity or intelligence to do that job, I have no idea where you got the idea I said that that’s not involved in that job… Nor do I have any clue how you took that away from what I was saying, is it opposite day today?
About a month ago I was doing a bare-bones games course and I like to do testing on my code early to find errors, so, there I was having an error I couldn’t solve and looking over the offending code for hours on end trying the find the answer, the code I typed in was exactly what was in the course.With a few minor tweaks as I was familiar with Godot.
So what to do?!? I asked a question on the forum of that site to the instructor, no reply. After two days, I was at wits end and decided to try and use the AI Assistant for an instant answer - I used the same question as I posed to the teacher, I wasn’t expecting it to be useful at all … and the AI assistant spat out an answer and told me how to correct the problem. I was shocked to say the least. Now I am using it a little more to try understand code that I can’t follow or explained by the teacher.
That’s great! But it also gives really bad advice, it’s not reliable or predictable, and especially: you need to actually know what you’re doing to assess whether it gives reasonable advice
And you also have LLMs that delete entire repos or databases and then apologize profusely and can’t explain why it did it
LLMs don’t think, they don’t reason, they guess words, some have some additional features connected to them allowing them to do some extra things but at the core they are nothing more than a far more advanced version of clicking the middle suggestion on your phone’s suggestion bar when typing, not in any meaningful way, that’s why they’re large language models
That’s not to say they can’t do impressive things, in the areas they are good at, but they need a lot of supervision and help to do really useful work, and are often highly likely to spit out utter nonsense, making them tools for people already knowing what they’re doing, not for people who can’t already do the thing in question or are able to verify the output. A game designer with no background or experience in programming has very low chance of making a game just using an LLM, because they wouldn’t be able to know how to fix the issues, or even identify them. Just like I couldn’t use an LLM to write something in Russian, without having someone confirming it’s actually valid Russian and not nonsense
It does, sometimes. That is the whole point. Sometimes.
I am fairly new to viewports (and they are amazing) but I was doing some 2D compass that had to infinitely scroll within a UI component indicating sectors like west3 or east4 etc and I had never done something like that. So I asked AI what would be a good approach and it said in 100% confidence that the best way was to use another viewport and to do the group chunked indicator ticks and labels I had expected and suggested, you know, so the last as it goes off screen becomes the first etc. I spent an entire day on it when I realised I could just use the ‘clip contents’ on a normal control node.
Anyway I asked AI if it would not be simpler to just do that and it said, ‘absolutely, that would be much simpler and is a normal approach to this in Godot.’.
Now thinking back, it probably suggested a viewport in the first place because I had been asking about viewports previously. But that was the wrong answer and a bad suggestion that, because I did not know, wasted hours for me. And also I felt stupid because I know you cannot take at face value what it suggests! And I fell for it anyway!
Yes it often helps. But also yes, it often does not help! That is why it is a dangerous tool for learning. Remind me of the format for this or that, fine. Take a look at this and what do you think, fine. For stuff you know, you know when it is telling you nonsense. If you do not know it in the first place, it is a dangerous guide that in the longer term, will waste more time than it saves.
As @athousandships said and others have constantly said here:
I almost spit my coffee out twice from laughing while reading this, this morning. Thanks @athousandships and @Demetrius_Dixon for that.
Also, I thought I’d mention that almost every interaction with @OriginalBadBoy on here ends up with us agreeing to disagree. Except that one time where we agreed on the solution to someone’s problem and we were both floored.
I’m using this in the future. It is such an accurate and amusing turn of phrase.
