conartist6

I don't think there's anything dishonest about being in the business of making tools. I am in that business. ...but I can only afford to fund that R&D because I first worked as the software equivalent of a plumber for many years.

Without the ability to own the things I build I never would have been motivated to build them, and without a career ladder I couldn't have worked my way up until I had the funds (and experience) needed to start my own venture.

What I mean by the term is creating value directly and in response to immediate needs. E.g. "your sink is broken > your sink is fixed" or "here is your burrito" or, ya know, taking care of kids and the elderly. It's stuff that needs no R&D, but also still has to be done for society to work.

Just out of curiosity, can anyone report on how the robots act in terms of right of way? Lets say the sidewalk narrows around an obstacle like a tree or trash can or lime scooter or whatever. If the robot approaches the constriction at the same time as an elderly pedestrian, how does it behave? Will it pause instead of entering? Will it back up a bit when a human would do so to be polite? Will it crowd in someone walking with a cane? In short, what are these things like to be around?

Yes. But when you act in protest, the goal is usually to gain attention.

Now the story is that the police came and protected an innocent robot by chaining up an old woman. That sure makes them seem like misanthropes.

"The shift in politics is even starker. In the 2010s and early 2020s, big tech was well-liked by most people. But AI is absolutely loathed by the public. These companies have never faced real public dissatisfaction, they all imagine themselves to be the good guys, with the credibility of having created magical search engines and smartphones. But that era is gone. They are now akin to the bankers on Wall Street in 2008, and politicians are winning elections by saying they hate AI as much as the voters do."

Yeah I see this too. That we're not innovating any more also doesn't help make our industry look good

I don't think this point is well stated. We have things like ECC to help with math being imperfect so that feels solvable. The problem with LLMs is much more subtle. They induce cognitive errors in humans. I'm doing it right now I'm saying "them" like they have individual identities, but they do not.

If you know they're just mirrors, then you know that looking in the mirror isn't the same as exploring the world.

It turns out, though, that many people can be trapped in a hall of mirrors. They will happily buy a condom for their brain.

I liquidated my retirement yesterday.

We all know the market is too good to be true.

We all know we created a great hollowness where once was our indomitable spirit. When a tree has died but still stands, what's left is the fall.

To the industry I once knew and loved: may the fall be swift, and may you be reborn from the ruins

The amount of trust AI has set on fire is incredible.

Business and people that I once though competent I trust no longer.

I'm watching leaders I thought had vision throw the future away for the shortest-term gains.

What have we become?

I wonder what you'll think of Paneditor.

It's like "what if the web browser became emacs".

I want to build it because it would put an experience with the level of power described so thoroughly here onto every device simultaneously. I think it would be an incredible boon for programming literacy.

At the same time we could make programming feel natural to do on touchscreen devices, and in languages other than English. Combined with just having a place to play and learn (a reason to write little programs) my hope is that such a tool could inspire a new generation of programmers who previously were locked out

Who is liable

I am building what you speak of!

You're saying there's nothing left to invent. I think that explains a lot of the complacency around considering AI to be the last ever frontier for progress. I obviously don't agree. I've gone looking into "why don't we have better version control" and the answer is definitely not that we already have all we could want.

It is progressing, yes, but it subsumed the productivity of humans.

Where before humans learned and built better tools, now you keep paying forever to get re-trained models. What makes it a regressive feedback loop is that the people who used to learn and push the frontier are now using AI to repackage others' work as their own. Eventually what will be left to train on but copies of copies of copies?