I saw Megalopolis and loved it. Mind you I went in knowing that it was apparently a $100 million Neil Breen flick. That only made me want to see it more though, since I've sat through all of Neil Breen's movies in awe of how bad they were.
Hard for me to call Megalopolis bad though.
It's operatic and impressionistic. People keep asking what it all means, I've never thought you had to know what a movie means to enjoy it.
Something like Southland Tales, meets Tree of Life, meets Neil Breen.
If you switch from sine to square or sawtooth the change in volume is jarring.
Waveforms having the same amplitude doesn't mean the ear hears it at the same volume. There's more frequencies contained in different waveforms and the ear is more sensitive to certain ranges.
A smoother representation of the waveforms might contain the first 5 - 10 frequencies of Fourier series but even then you need to compensate with a change in amplitude.
It's not a DAW, it's a web app, and the users will appreciate the favor.
I would like to have a more modular approach with more specialized training of the models.
Currently I can only really use smaller models for micro-tasks like sentiment analysis or classification, but any type of problem solving has to be left to GPT-4.
With these examples I think the author would still be stuck with stepping through the state machines with the students. Unless what you wrote would allow for the "autonomousPeriodic function to keep ticking" another way?
Apologies for not making that more explicit. My point was that that isn't necessarily code, but also data. Literally, you can turn that into a list and instead of evaluating it with the standard runtime, you can send it to another place that turns it into the "command" style from the example java.
You can /kind/ of do this with java, of course. Just make sure to not use "new FooCommand" and instead change the "foo" function to return a the command object. No reason that couldn't be done; but, and this is the big difference, it requires building a ton of scaffolding in the java program to support both ideas at the same time. In lisp, it is fairly easy to wrap in a macro. Still somewhat magical, I suppose, but no more so than the rest of the compilation/build process.
That make sense? I'm somewhat interested in this, so more than happy to try and do a blog post on the idea, if that would help.
Also, how is typing the questions in an email to a grad student simpler than using the chatGPT UI. He's instructed his own authentic intelligence assistant to interact with the artificial intelligence assistant for him.
I don't think it's a solution looking for a problem. I'd be willing to put them on at work and see which pins on a piece of hardware do what instead of looking back and forth between a datasheet. Lots of examples exactly like that, especially if the glasses are fed sensor data so the temperature/pressure appears right beside the area it is measuring.
Like I said, industrial environments. A small niche group of people and businesses who could use some information in front of their face as they do hands-on work.
How is this feasible at all? Not only all the tech for the AR device to be light and have decent battery life, but the software you describe would need optical recognition of what you're looking at, then match that to the manual of the object to tell you what the pins do. All on a device with significant size and heat constraints.
For early generations I'd imagine you'd have to plug them in via usb (which would still be fine for my use case I was describing). Processing done on a different device, get some type of AI to summarize datasheets and match it to pins. All of which probably doesn't sound good to you but I'd still prefer it over turning my neck and scrolling back and forth.
But like the parent poster pointed out, I'm now talking about niche industrial uses, not widescale adoption.
My thoughts exactly - rewrite the novel with Mr. Carraway as an ML engineer while maintaining themes/motifs (possible adding new ones too). I'm guessing what's impressive is that these are the first steps towards something like this? Or is it already possible? Someone please correct me here.
Or rewrite The Count of Monte Cristo as a science fiction novel and get The Stars My Destination. Or rewrite Heinlein's Double Star into present-day and get the movie Dave.
I don't know if it's so cut-and-dry. By that logic why doesn't everybody just do amphetamines all day.
Probably some integral of pleasure is part of the equation, and it depends on you preference for pleasure distribution. But maybe I'm blurring the line between pleasure and fulfillment.
The brain is made up of modules and happiness in essence comes from what can be characterized as two different modules. For alignment both modules need to register happiness.
The animistic or instinctual side of happiness triggered by drugs is not purely a module but it can be encapsulated into a concept called the the mesolimbic dopamine pathway. This pathway can be mechanically manipulated to make you feel a certain portion of happiness via drugs sex or the natural secretion of dopamine.
The thing about this module is that it lives separately from another module called the neocortex. The neocortex encompasses the higher functionality of your brain with things like higher order thoughts, perception and consciousness. The unique thing about the neocortex is that it can self actualize. It can see signals being sent by other parts of the brain and it can override or interpret those signals as false. This is what happens when you do something like exercise. One part of your brain tells you you're tired, but your neocortex overrides this and forces you to continue because you're aware that the feeling is an illusion. Exercise is actually healthy.
For someone to feel truly happy the mesolimbic dopamine pathway and the neocortex need to be in alignment. The dopamine hit must register AND the neocortex must validate that feeling as well. Signals of happiness from one module but not the other register as sort of an invalid false happiness.
For example when you're an addict and you're doing heroin the reward pathway sends a signal, but the neocortex denies it. The addict is generally unhappy even though he gets "high" all the time. For the parent poster, the neocortex is telling himself he's happy by "thinking" but there's no dopamine hit from the mesolimbic pathway. The in-congruence from both modules in the brain form a happiness that isn't quite real. To truly be happy you need both.
Also note that the brain additionally develops tolerance to dopamine so any feeling of happiness is relatively fleeting. You need to use different stimuli and go through long periods of "no dopamine" to feel happy. Humans are biologically designed this way by evolution.
I'm sorry to tell you this, because in general being aware of all of this generally makes people less happy. Ignorance is part of what makes all of this work, because it is your neocortex that is determining all of this. Knowledge of this influences how your neocortex chooses to override certain signals from the mesolimbic pathway... it's all very very meta.
No need to be sorry, I'm aware we answer to our biology (not that I have as deep of an understanding of it as you).
We may be splitting hairs because this stemmed from a comment saying thinking is the most pleasurable thing in the world and you said anyone who thinks that way is lying to themselves.
Through your biological lens I have no choice but to agree, the thing that is most pleasurable in the world is mesolimbic dopamine pathway and the neocortex alignment (or so I'm told).
But through a conversational lens I guess it's all how you get there. And thinking->problem solving->success is a plausible path to get there which is why I disagree with your statement that they are lying to themselves, ignoring the existential truth that we are always lying to ourselves.
>And thinking->problem solving->success is a plausible path to get there
Sure, if this is what he means, then he is correct. But what he wrote is just "thinking" which is false.
>which is why I disagree with your statement that they are lying to themselves, ignoring the existential truth that we are always lying to ourselves.
It is a bit. The success of problem solving usually comes with only a mild dopamine hit. There are way more situations that create a greater surge of dopamine. Parent stated "thinking" was the most pleasurable thing ever which obviously not the case for people in general.
For example it's much more pleasurable to win a billion dollars or successfully bed the hottest girl in the club or to dominate everyone else in some competition. Clearly our biology is geared towards giving huge rushes of dopamine for situations that aren't exactly tied to "thinking" or strictly "problem solving". I find it hard to believe that the parent is completely unaware of this. He must know on some level. That's why I claim he's lying to himself.
Hard for me to call Megalopolis bad though.
It's operatic and impressionistic. People keep asking what it all means, I've never thought you had to know what a movie means to enjoy it.
Something like Southland Tales, meets Tree of Life, meets Neil Breen.