trention

>Nobody can stop the march of progress.

Several periods of known human history, starting from the late bronze age, disagree.

>It's hard to imagine the mental gymnastics required to think the two are in the same league.

It's really not hard. You just have to care more for the harm to the well-being of human artists than for the harm for the well-being of farmed animals. It may surprise some people but speciesism is the default position for the vast majority of the population.

What you're describing is essentially a DSL. We can ruminate on the exact level of details that will be needed but in my opinion by necessity it will be a lot higher than what most commenters in those threads are imagining. Open API specs are already quite complex and verbose and it's not exactly clear to me we can do a lot better when describing APIs, whether web ones or not.

As I said, if I can't distinguish it, I will stop reading "it" at all. There is no utility in reading text that is not human-generated.

>off the prompt "like Facebook only better"

This prompt is unspecific to the point of unusableness. Even if this works some day, the spec used will be a lot more detailed, in higher-level pseudocode style.

Diminishing returns.

Why would I want to read a blog post written by an AI about a program written by an AI besides novelty? The moment an input in a medium becomes primarily AI-generated is the moment I stop using said medium.

>People will learn to pay more attention to the reputation and publishers will apply more scrutiny in return

A subset of the people will but most will not. Right now obviously fake news stories are shared left and right, without even a modicum of rational analysis applied as to their plausibility.

The average person is simply incapable of rational thinking about anything that is not day-to-day and this will not change.

My personal opinion is that it's unethical (and possibly illegal, in a subset of cases) to train models on data without explicit consent of the creators of that data. And that really encompasses all data - generative models were not a thing when said data was created and no matter how it was licensed before, explicit consent about using it for model training must be obtained from the creators themselves.

That being said, arguments about copyright are just a fig leaf as far as I am concerned. The outcome of whether this is allowed or not will depend on the net impact of using those models on the job market and whether society will be willing to tolerate it.

If you are well-aware, why didn't you use a more adequate money supply measure like M2 or M3? Is it maybe because the growth there was nothing out of the ordinary looking at the past 10 years?

And I think your hunch is about as correct as your rationale for posting the M1 graphic without reading the note below it or noticing that it jumped 4-fold in a single month.

If you think you will be able to tell by looking at a period of 6 months whether a contraction (that hasn't actually even started yet in terms of total employment) is just a regular recession or the apocalypse you prophesize, I have a brooklyn bridge to sell you.

There aren't "hundreds of millions" of developers on this planet. The talent pool is small because just a small fraction of the workforce worldwide has the training, skills and capabilities to work as software engineers.

There is no scenario in which tech declines without the US entering a recession and when the US enters a recession, the "cheap money" is going to be coming back.

There is no way the dotcom crash is repeated because today IT is integrated inside every aspect of every business. It will be just another regular recession.

The biggest threat to programming jobs in the US is outsourcing (and, in the longer term - automation), not an imaginary apocalyptic event.

Because the AI is not a person and there is no reason to treat it like one, including by allowing it to "learn by looking at other peoples work".

Or maybe I haven't and GPT-4 will be ~on par with PaLM, already exhibiting diminishing returns as expected.

>cannot generate coherent dialog (and even then it's mostly bad dialog) spanning more than a minute or two

I think that was pretty clear and that posted dialog is a perfect illustration.

You cannot generate the entire movie script coherently without significant human input and that's not going to change in the next several years. So, your initial claim that dialogue is "solved" is indeed false.

>then the real thing happens earlier than the most optimistic dates

Is this like with self-driving cars that were supposed to be a consumer product in 5 years...in 2012?

Or like when Hinton said radiologists will be completely replaced in 5 years...in 2016?

By the way, Moore's law hasn't been a thing for a while in its original spirit.

>Characters and dialogue are effectively solved, just look at GPT-3.

Is this the motherload of exaggeration?

Current language models cannot generate coherent dialog (and even then it's mostly bad dialog) spanning more than a minute or two. And their current capabilities in that area are definitely significantly below those of the average human writer.

There weren't "uprisings" mostly because the destroyed jobs were replaced with others that paid better and were less "dangerous". There were some local minima (Detroit auto workers) where this happen only partially and we know the pathology this led to.

Is this replacement of jobs happening this time around? No. So violence there will be.

Think of that 1 byte as a hyper-compressed essence of knowledge necessary to make said collage. Then it will be correct. Those models would not exist without the labor of the artist community - which in this case was used without consent and without pay.

Scale and access (≈quantity) matter.

There is enough work already done in game theory about the dangers of nuclear escalation. Current state of affairs pretty much ensures no rational agent will want to use them. On the other hand, when you model the interactions of significantly more intellectually capable rational agents with us, the "calculus" is not very favorable, to put it mildly.

Short-term, textile workers in the UK at the time were definitely harmed by automation. They eventually managed to move to better paying jobs because those were available. There is no reason to believe there will be any jobs left in the age of AGI. And since the default state of humanity is existence at the subsistence level...

>that’s a gain for humanity

Imagine what has happened to american manufacturing workers since the 80s but happening to everyone on the planet. This is by no means a "gain".

>Has he ever been drastically wrong

Has he ever made a prediction with as drastic a consequence and transformative potential as the one he is currently making about AGI? Actually, is there any track record of predictions he has made and how they have panned out?

The old problem of induction. It's interesting how many of the AI accelerationist crowd are (purposefully pretending to be) unaware of it.

That scaling will eventually hit a wall. What was it about nerds and S-curves?

AGI will be more dangerous that nuclear weapons.

People are not allowed to start a nuclear weapon company. At all.

Why are people allowed to casually start an AGI company?

Because the history of the species on this planet clearly indicates that the smartest one will brutalize and exploit all the rest. There are good economic (and just plainly logical) reasons why adding "artificial" to the equation will not change that.