hsm3

How much of this is at all related to newer Gen AI abilities and LLMs, vs just more/better "traditional" ML?

Do have have any links handy to those test results?

Honest question: Does the kid really want learn via dealing with a book? Programming has such a wonderful virtuous feelback look built in, seems a shame to get in the way of that. My son learned a ton of programming mainly by building bigger and bigger things (with more and less help from me), and only after many years got to a point where researching a question on stackoverflow/etc started working for him. All the progress was well-motivated by the task at hand, not the next page of the book.

One tool that we found to be a very deep learning ground: an iPad app called Tynker. It is another block-based env like Scratch, though we found Tynker to have stronger primitives (e.g., you wind up forced to use globals to pass state around less often). Some big advantages: 1) iPad is super portable, so works when travelling 2) has an excellent physics engine built in - we made an Angry Birds clone 3) was strong enough to support making a piano app that can play concurrent tones. 4) Lots of samples shared by community for inspiration.

You might enjoy this documentary from Telluride Film Festival last year https://www.imdb.com/title/tt21956346/

NeXT wasn't bought until late 1996. That didnt kill the RISC project, what happened is the company dropped all hardware and became software-only (primarily on top of x86 boxes, but some lower-level pieces ran on HP and SPARC).

The original OS absolutely had shared libraries - the whole environment was so RAM-constrained that it was essential. (source - I was part of the early SW eng team). In fact, my memory is that they worked much better than competing systems because of Objective-C's late binding, in the practical case of sharing library code across applications built/shipped at different times on different OS versions. As I recall, C++ based systems couldn't deliver the full sharing on a typical user's system who was never running apps that were all updated for each major and minor release.

I feel like the Isaacson bio is fairly accurate factually on that time, from what I knew as a senior SW Eng at NeXT/Apple. It was a huge course change, eventually pushed through the entire software org by Steve, the NeXT leadership and a set of Apple folks that chose to embrace the new direction.

I strongly disagree. Amelio's Macworld 1997 keynote was a legendary failure, just a shocking public show of incompetence. He built zero confidence in anyone joining from NeXT. His book shows a poor understanding of what happened at Apple ("wow, I directed a VP to go write a new modern OS, and they just never did it!?!"). There was no useful depth of vision or insight coming from him or Hancock at the time (my opinion as a lead eng from that time), and it is clear to me that Apple would have continued to spiral into oblivion under their lead (like yahoo).

I'll grant that he did spark the revival by doing the NeXT deal, as Apple did need an acquisition to reboot the software stack.

Gifted kids are all really different in their specific strengths. That said, for one with a well-expressed nerd gene, we started at age 6. I would start with an iPad app (slightly simpler container than a browser), maybe that gamifies the whole thing a bit with some easy on-ramps. For us that was Tynker and a bit of Hopscotch. It's pretty easy to tell whether you've got any sparks of interest on your hands, else no point in pushing it.

My son and I got a lot out of a similar iPad app called Tynker, which we both ended up liking better than Scratch, which he used at school. Tynker has a physics engine, which opens up all sort of great game possibilities (we made a cheapo Angry Birds). Also as referenced above, Tynker has a slightly richer language that gets rid of the need for some bad practices (e.g., functions can return values, you can declare local vars vs all-globals-all-the-time, etc). Having the tool work well on a tablet also travels really well, and is a simpler container than a browser window for a really young kid.

We've not had a big problem getting candidate streams of people who want to work with Ruby. In a way I don't mind having an implicit filter against people who only want to work with shiny new things, as we don't really want our platform filled with everyone's favorite shiny technologies any more than we need a shiny language to write a good-sized e-commerce web app.

https://en.wikipedia.org/wiki/Great_conjunction gives the exact time as 21 December 2020, 13:22:00 UTC, or 5:22 AM PST. In SF, the planets will set around 7:10 PM, the sun sets about 5pm. So if my best observation time will be around 5:30 or 6pm, seems like they will be at least as close if not better on the 20th as the 21st, right? And then even more so for HI, two hours later.

This might work for some people, but for me it would have been a worthless and tragically misleading signal. I got modest-at-best enjoyment from being around kids before I had one, but the past 10 years with our son has held many of the best experiences of my life. Your own kid is a very different thing than all the other kids you might spend time with.

My experience is more like pg describes, where I had a lot of trouble projecting what it would be like, and in retrospect all the rational pro/con analysis I did wasn't really getting to the core of the matter.

Aren't most of those shenanigans done early in the VM's lifespan (as Rails boots, actions are called the first time,...), after which the world would be stable, and JIT-related caches wouldn't have to be lost?

I was leading the AppKit group at NeXT at the time, and "NS" definitely came before the Sun relationship, referring to NextStep (or NeXTSTEP, the spelling changed every release). I distinctly remember our team expecting we'd have to suck it up and change the prefix to something more neutral when the Sun deal was made, and then being surprised that they didn't seem to care when the spec was discussed. Maybe the alternate interpretation of "NeXT and Sun" satisfied them, but I don't remember that expression ever being commonly used, and it was not the original basis for NS.