Libraries. I use both languages, and a survey of what libraries are available is part of picking an implementation language when starting a greenfield project.
I went the other way: from a Pixelbook to a Macbook Air. I mostly do SW development in the CLI, so the Linux subsystem on the chromebook was fine, as is macports/homebrew/etc. on the mac. I would still be using the Pixelbook if I could have replaced its battery. The low-end Air had good price-performance tradeoff, and the Neo would probably be today's choice.
The Greenland shark appears, if I remember correctly, in her book "Golden Mole", which is about many interesting creatures. This is published as Vanishing Treasures in some countries. Her "Super-Infinite: The Transformations of John Donne" is interesting and also not a children's book.
I enjoyed Werner Herzog's "Encounters at the End of the World" at many levels, not the least of which was how different it was from "Aguirre, the Wrath of God".
I used to ask myself the same question, but then I realized that for these people it doesn't matter how much they spend. When you are worth billions of dollars, the difference between spending $10M or $50M on your home Does Not Matter. You still have many other $M to spend on other things. It's perfectly rational for them to spend what seems like a large amount of money for an apparently small marginal improvement.
"You can learn anything now. I mean anything." This was true before before LLMs. What's changed is how much work it is to get an "answer". If the LLM hands you that answer, you've foregone learning that you might otherwise have gotten by (painfully) working out the answer yourself. There is a trade-off: getting an answer now versus learning for the future. I recently used an LLM to translate a Linux program to Windows because I wanted the program Right Now and decided that was more important than learning those Windows APIs. But I did give up a learning opportunity.
There's a mystique around Mathematica's math engine. Is this groundless, or will you eventually run into problems getting correct, identical answers -- especially for answers that Mathematic derives symbolically? The capabilities and results of the computer algebra systems that I've used varied widely.
The surge in laptops contributed, too. The opportunity or need for expansion cards, additional memory or storage upgrades, and peripherals disappeared or shrank.
I used to think of the sales staff as the United Nations of Fry's. It was always thrilling to see someone starting their American dream, even if the service was haphazard.
I was at G when "mobile first" was the slogan, and it led to "odd" choices such as designing and leading with a travel app rather than the web site. Perhaps locally suboptimal, but in the long run brutal forcing functions were needed to move a company as big and successful as Google into something new. I hear that going all-in on AI was internally disruptive and probably had some bad side-effects that I'm ignoring, but in hindsight it was the right thing to do. When ChatGPT, perplexity, and you.com came out, my immediate thought was "Google is toast", but they've recovered.
Does anyone know what the "30,000 star trackers" are? There are about that many satellites in orbit, so does this mean almost everyone is contributing data?
Mercator projection aside, Greenland is not tiny. The ice-free area is about as big as California. If you include the icy areas, it's bigger than any U.S. state.
This is looking at the wrong end of the telescope. The arc has been to move computing closer to more and more end users. In the 1960's, FORTRAN enabled scientists and engineers to implement solutions without knowing much about the underlying computer. Thompson and Ritchie got a PDP11 by promising to make a text processing system for patent applications. Many years later desktop PC's and programs like VisiCalc and PageMaker opened up computing to many more users. The list goes on and on. With this movement, developer jobs disappeared or changed.
I once had an 1.75 hour commute each way, 3.5 hours daily. I had to program during the commute---otherwise there weren't enough hours in the day to get my work done. There were periods of laptop-closed thinking, but no daydreaming or looking at the scenery. No Internet connectivity, so I had to plan carefully.
Can anyone provide some color around this: "I started porting esbuild's JSX & TypeScript transpiler from Go to Zig"? Hypothetical benefits include monolanguage for development, better interoperability with C and C++, no garbage collection, and better performance. What turned out to be realized and relevant here? Please, no speculation or language flames or wars.
The AlphaGo paper might be what you need. It requires some work to understand, but is clearly written. I read it when it came out and was confident enough to give a talk on it. (I don't have the slides any more; I did this when I was at a FAANG and left them behind.)
Three considerations that come into play in deciding about using RL: 1) how informative is the loss on each example, 2) can you see how to adjust the model based on the loss signal, and 3) how complex is the feature space?
For the house value problem, you can quantify how far the prediction is from the true value, there are lots of regression models with proven methods of adjusting the model parameters (e.g. gradient descent), and the feature space comprises mostly monotone, weakly interacting features like quality of neighborhood schools and square footage. It's a "traditional" problem and can be solved as well as possible by the traditional methods we know and love. RL is unnecessary, might require more data than you have, and might produce an inferior result.
In contrast, for a sequential decision problem like playing go, the binary won-lost signal doesn't tell us much about how well or poorly the game was played, it's not clear how to improve the strategy, and there are a large number of moves at each turn with no evident ranking. In this setting RL is a difficult but possible approach.