figure8

You do realize that that's also likely true for all the thousands of Gods that've ever been imagined by humans? So I am gently nudging you to consider praying to them as well, just to be on the safe side. ;)

Those are some awesome videos. You'll notice that by the end of the series, he's using advanced notation everywhere as well. And you'll notice the original PDF has lots of pictures too. The use of calculus notation is required to be precise about differential geometry. It is not some kind of elitism, it's just the most effective way to communicate abstract ideas without misunderstanding. I hit a wall in my math understanding until I was willing to embrace the symbols, rules and manipulations of good notation.

I'd go further and say that mathematical and physics notation is beautiful unto itself. The evolution of symbols used to transmit abstraction is incredible. They are way of exploring patterns which can then be refined and layered using proven notational magic. So cool.

> But can anybody name a wrapper company as/more valuable than foundation model companies

Ask the same question about databases. Then consider these wrapper companies: Salesforce, LinkedIn, AirBnB, Workday, ...

The only database company worth more then Salesforce is Oracle. Other database companies are worth much less.

I have a very naive question.

Why is the ARC challenge difficult but coding problems are easy? The two examples they give for ARC (border width and square filling) are much simpler than pattern awareness I see simple models find in code everyday.

What am I misunderstanding? Is it that one is a visual grid context which is unfamiliar?

I've always thought that using vertical space and indentation alone to organize code was a failure of imagination for IDE development. Why can't the IDE allow for and auto-format code in a horizontal line? For instance, I can easily imagine background color shading as an indicator of logical depth. That would be fun to play with.

And along those lines, why do coding best practices always encourage vertically stretched out logic? Its analogous to writing all literature as poetry. That's clearly suboptimal or at least limiting if you can read multiple ideas in a horizontal line.

So, I think, ideally, semantically-aware IDEs should allow for any formating and supply dynamic adjusters which can reformat code to the users preference, then save it in a presentation-agnostic encoding.

Yes. I appreciate the article because it gets people to think about high dimensional geometry. But I think the right takeaway is that in high dimensions, cubes have huge volume compared to the enclosed sphere. So the sphere is not spikey at all. Instead realize that the diagonals and volume of the cube get very large compared to the sphere.

I subscribe to YouTube Premium and strongly recommend it. With my heavy usage, I'm happy to pay each month to avoid ads.

Latest from their status page: Update - The underlying issue affecting a subset of our API customers (and ChatGPT by extension) is with an upstream part of our API authentication stack. We're simultaneously scaling up the underlying cache resource we found to be near capacity, and working on an alternate system that does not use the cache. We will keep you posted as soon as we have an update, and are sorry for the trouble this is causing you. Oct 19, 2023 - 19:27 PDT

Try using a mobile browser, like Opera, that supports text rewrap on zoom. It is a revelation.

This depends crucially on your definition of AGI. LLMs are more like a mathematical function than they are a conscious being. If, in your opinion, AGI could be realized as a input/output function with no changing internal state, like a lens or a lookup table, then we could say LLMs could be an element of AGI. But if, like many of us, you believe an AGI needs a changing internal representation of the world, and an ability to mull over prior knowledge and incorporate new inputs, then LLMs are at best only a useful component of AGI. Maybe like the retina of the human eye plays some role in human visual intelligence.

The fact that LLMs appear so intelligent to humans is really a reflection of our inability to imagine effects of scale. We can understand simple linear predictions as trivial calculations, but when language-based pattern discovery is many layers deep and those patterns are combined in nontrivial (but non-intelligent) ways, we project intelligence onto the result.

I agree. This is why I use Opera's mobile browser which defaults to zoom with dynamic text wrapping to fit the screen width.