So what you're saying is that there's plenty of good things about HTML. Just that no one's really using it?

Coming from a gaming background, the web seems awesome. I can publish my game without gatekeepers like Apple or Steam. But, building the game itself using HTML/JS/WebAssembly has too many limitations.

That the whole stack is bloated to support various use cases, how can that be solved differently?

Would you stick with HTML if you'd write a 'new browser' from scratch?

I'd stick with pure JS for web apps, and markup (HTML and evolutions thereof for the content we're actually creating) for documents with basic GUIs such as tabs, menus, and the like.

This sounds like the issue where a 'datepicker' is not implemented in a standard way across browsers.

Like each browser does it differently, and on top of that renders emojis differently. Is this similar to the issue you're talking about?

It does feel like 'hacking' quite often to get special UIs done in HTML.

What do you think is a better platform for the web?

That meets similar requirements: - fast to deploy/load cross platform - accessible to build with

There is no serious alternative for the purely editorial stuff but still lots of room for improvement.

For the application-like stuff… there was some good and some bad in Java applets, Flex, and Silverlight but none of that had a barrier as low as HTML/CSS/JS.

I think the way forward is for HTML and the DOM is to adopt ideas from the framework space (like one/two-way binding) and expose more high-level UI elements out of the box.

Just imagine something like:

    <script>
    const myDataTable = new DataTable()
    myDataTable.head = ["Foo", "Bar", "Baz"]
    myDataTable.items = [[1, 2, 3],[4, 5, 6]]
    </script>
    <table binding="myDataTable" class="my-data-table" sortable />

with everything needed built into the platform, and no hard need for any third-party framework.

Having never used XML + XSLT I can't understand exactly what that means.

Do you imagine a better version of HTML to be more like that? Or perhaps like SwiftUI? If you're familiar with that.

So because coding art is hard, it's more difficult not to make it minimal, I guess.

Is that built with something like 3djs?

Built using my custom creative coding tool which also incorporates some d3 stuff:

Code sample: https://workbench.hyperobjects.design/share/?script=5febbf46...

If you'd be interested, you might want to join a beta test for a collaborative design for iPad: https://www.apance.com

This is all the stuff launched by Ariane rockets, or is it related to a single space program?

Thanks for the question! These are all Ariane leftover rocket parts, related to various launches. The leftover parts are mostly booster stages to transfer a sats to GEO orbit.

Any chance I could find out about satellites flying above my home town through this? Perhaps a quick help guide could help explain how it works.

It would be cool to see what kind of sensor data is available from my garden.

Yeah should be possible to add a feature like this in the near future! Will add it to the roadmap :)

How do you see this specifically creating AGI?

How would you calculate the loss on a reinforcement learning model that's autonomous?

Allegory:

Running entity A performs action at time t_sub(0) that costs n_sub(t_sub(0)) currency. Response to this action by the counterparty creates a cascading tree of potential new actions, each of which require individual “reconciliations” (new actions, which each require a new response), defined by the probability distribution of potential responses->new actions. We either know these distributions a priori based on our initial conditions, or we can create them based on an initialization function.

The net present value of the expected action to these new responses can be evaluated with respect to the NPV of the current holdings of the running entity’s portfolio, and that difference can be treated as the loss function.

I’m not an ML researcher, so I apologize if my lack of terminology makes this sound stupid to you but that’s my initial thinking.

Feel free to email me if you’d like to discuss further, I’ve been tangentially working in this area for a while but this really gets my sparkplugs going.

I don't think trying to shove this model into a gradient-descent framework makes the most sense here. I'm an ML (industry) researcher and I highly doubt that AGI will be achieved with gradient descent on neural networks alone. Those may play a small role somewhere in the stack but the orchestration and reasoning will be managed by something else entirely. Neural networks today are fancy MLE machines -- nowhere close to reasoning machines, which require an "understanding" (whatever that means in this context) of dynamics with respect to the environment.

Seems more appropriate to start with a population of agents who reproduce at a rate proportional to their recent rewards, and allow them to die off at a rate inversely proportional to the same, a la a continuously-evolving genetic algorithm setup. You may have to modify the reward function to disincentivize behaviors which cause systemwide collapse, but that goes without saying.

> You may have to modify the reward function to disincentivize behaviors which cause systemwide collapse, but that goes without saying.

Hopefully humans will figure this out one day too.