Indeed. If the DNC had allowed Bernie Sanders to run against Trump, Sanders would very likely have beaten him and Trump would have remained a reality TV star. Instead, they pushed Clinton on Democrats just like they pushed Harris this cycle. The corruption within the DNC is very much to blame for Trump.
I refuse to accept the childish assertion that the majority of American voters are card carrying members of a radical political system that was defeated last century. This is just an emotional response to a reality that one does not want to accept.
I recently just started using psst which is a Spotify GUI that's much lighter. When you right click a song and go to show similar tracks u get an array of sliders corresponding to the audio analysis/features like valence, danceability, energy, etc to tweak the recommendations.
It made a light and day difference for music discoverability for me, while the default spotify radio keeps giving me songs i skip instantly multiple times along with songs I've listened to a hundred times, doing this through the API, is 100x better. I've discovered 30 new songs that I love this past week while that number has been steadily dwindling for the past 6 months using Spotify.
> key problem in both cases is imposing the requirement to be connected to the internet, after which it's game over
My point is we can do this for anything. Clocks enabled capitalists to exploit workers. Lights took the night from rebels. Steam engines took farmers out of the daylight and put them in mechanised factories. Hell, agriculture enabled modern civilisation but also created such a wave of oppression and anxiety we're still worshiping the gods we invented to save us from it.
None of those statements is wrong. But they're also incomplete.
I think vacuum energy is best described as a product of the mathematical model used to describe reality with an accuracy of 10 decimal places. From that perspective it seems a very reasonable deduction.
Wave functions can be seen as the sum of polynomial terms.
So a wave function of: x^3+3x^2+4x could be expressed as its coefficients [1,3,4].
Now if any of those coefficients are allowed to be zero, then you’d have, say, [0,3,4] for your polynomial which simplifies to an identical polynomial [3,4].
This just reduced the dimension of the problem by 1, but a wave function in n dimensions is fundamentally different than one with n-1 or n+1.
If you could reduce the dimension of a problem by zeroing a coefficient and still get accurate calculations then then “curse of dimensionality” would be moot because you could just keep reducing the dimension until the problem was tractable and then reverse the process to build back up to the original problem.
Unfortunately, the pigeon hole problem creeps up when the additional states of the higher dimension lack a direct analogue to its lower dimension.
Think of a binary tree. At each element draw two new elements branching from it on a line below. Allowing any leaf to be 0 effectively removes it and all of its parents and children from the model. So in order to retain the full tree all values at each leaf must be greater than 0.
RMarkdown in RStudio was the killer feature, until the VSCode R extension matured. Not only does it support RMarkdown, it adds a ton of features RStudio doesn't have and runs a lot faster. https://github.com/REditorSupport/vscode-R/wiki/R-Markdown
For my uses, it replaced RStudio 100% of the time.
LEGO is indeed pretty expensive these times. Not every set is that bad, but the most cars are not only expensive, but not even the best sets on the market...
There is a lot of comparisons, but the most convincing one maybe the LEGO Ferarri (42125) vs. CADA Italian Super Car (C61042W) video[1] of "Held der Steine" (german youtube channel). The LEGO is "parts only" - the CADA is remote controlled with led lights near double the parts of the lego one for the same price.
Many of the newer manufacturers are 100% compatible and provide near double the part count for a similar set. Even if some of them are shameless rip-offs and illegal to have in some countries, most of them are smaller companies with great part quality, prints (no stickers) and innovative designs (see CaDA® C61503W AMG One).
There is also the world of Trains or Tractors, where LEGO tends to fail every time.
While I agree with your general premise, I'm not convinced that remote work is the driver we should rely on. As a peer comment mentioned, infrastructure is a weak point right now, and remote work would add another dynamic that would make the situation more risky instead of less. Up front though, I should clarify I am a huge proponent FOR remote work. The rest of this is focused on why remote work should not be a foundation to build American housing on.
Prior to the normalizing of remote work and cloud computing, the infrastructure risks that a company needed to consider were related to hubs. Cloud computing moved a lot of the processing out of the hub, which is good from a risk perspective. This leaves the need to ensure the infrastructure related to workers accessing the computing is resilient.
If remote work becomes the foundation we building our cities on, we now expose our companies to the additional infrastructure problem related to internet connectivity while not resolving the connectivity issues inherent in the other grids of roads, power, and water. This is fine for companies that are natively born to this, but this is dangerous for the existing large cap companies and governments that are not.
And just as we're pushing for an increase in remote work, we are also in a period of time where our infrastructure is regularly attacked remotely.
Again, this is not to say remote work is bad. There's just a lot of transformation that needs to occur and I personally feel we should not take a darwinian approach to this when state and local governments are involved.
The instability we would introduce through this would likely lead to corporate funded infrastructure being stood up to ensure remote workers maintain access to cloud computing. I believe the company town concept [0] would make a comeback.
My main fear is that we would inadvertently create remote private corporate fiefdoms that would lead to corporate scrip [1] being used for local goods and services and non-transferable to other regions. The flexibility of remote work would, if my fears are realized, lead to a world of less flexibility than we have today. Not more.
I don't know what a better driver is though. How does one generate desire for a traditionally undesireable ___location?
If you're talking about instability / outages as the sibling comment is, can't help there, but if you're running into laziness (I never use the standard one for coding) i spent a while on this custom GPT and it works well.
If you have at least some coding experience and you are interested in the practical aspects of ML/DL (i.e., you want to learn the how-to, not the why or the whence), my recommendation is to start with the fast.ai courses by Jeremy Howard (co-author of this "Matrix Calculus" cheat sheet) and Rachel Thomas[a]:
The fast.ai courses spend very little time on theory, and you can follow the videos at your own pace.
Books:
* The best books on ML (excluding DL), in my view, are "An Introduction to Statistical Learning" by James, Witten, Hastie and Tibshirani, and "The Elements of Statistical Learning" by Hastie, Tibshirani and Friedman. The Elements arguably belongs on every ML practitioner's bookshelf -- it's a fantastic reference manual.[b]
* The only book on DL that I'm aware of is "Deep Learning," by Goodfellow, Bengio and Courville. It's a good book, but I suggest holding off on reading it until you've had a chance to experiment with a range of deep learning models. Otherwise, you will get very little useful out of it.[c]
Some sections may be less relevant, depending on what you want to do, but Section III is a very good introduction to machine learning methods.
Do the exercises as you're reading. Theory is one thing, but in ML my rule of thumb is that you don't really understand a model until you've coded it up. A collection of written exercises would be a good way to impress an interviewer, too.
One can make a very good argument that in elementary calculus, we actually use what logicians would call "terms", but refer to them as "functions". For example, if f is a unary function symbol and x is a variable then f(x) is a term, different from f(y) if y is a different variable. It's possible to develop everything quite rigorously using this machinery and when the dust clears, you get a rigorous version of what is actually done in practice in the elementary calculus classroom. As an advantage, certain things become much clearer, for instance, there's a very nice abstract multivariable chain rule which I describe in this paper: https://philpapers.org/archive/ALEFDV.pdf
I think Huffman is falling into a trap that Kanye and Mark Zuckerberg have fallen into: trying to prove he's special.
Kanye rose to fame as a young genius rapper/producer/designer and was acclaimed repeatedly. Zuckerberg rose to fame as a brilliant business twenty-something (according to the world around).
They spent half their life being praised and told they're important, brilliant, special, prescient, etc. So they are confident they can prove it again.
Kanye is doing.... something strange now. Zuckerberg tanked the stock price with VR/Metaverse and then juked it by firing a chunk of the company. They are so confident something intrinsic in them makes their choices special and 'right,' that they can see something no one else can.
Huffman et al. were at the right place and right time with a LAMP(?) stack when Digg imploded. Now he wants to prove that it's not "right place right time" but that he is special.
I refuse to accept the childish assertion that the majority of American voters are card carrying members of a radical political system that was defeated last century. This is just an emotional response to a reality that one does not want to accept.