suref

An HTML file is all you need.

You should try Midjourney.

That’s an implication I haven’t really thought about, but it does make sense. It’s a lot harder to fix all the problems in the world than choosing actions for yourself only though.

It’s fanatstic how quickly these models can be shared with anyone.

Is it smooth? Often you would like to add it with just one or two clicks, like you do natively on each app. Otherwise it's a bit annoying.

I like it, but only for twitter.

I have different reading lists on hacker news, twitter, reddit and medium and because of this I never read anything that I don’t read directly… If you need to share between them you need some convenient app for your phone and computer.

Interesting take. There are definitely multi-objective optimizations at play and as you say knowing this should be advantageous. As opposed to running it in code there is more uncertainty, partial information, less options for tests/objective evaluation etc. though. In fact we as individuals are doing multi-objective optimization every day when we spend our time to achieve different objectives (staying healthy, earning money, having fun etc.).

Alright, I see your point. My idea was to create a second article or part 2 with some more practical work using MOGAs if there was any interest. But I can see the benefit of adding a simple example here too.

There are plenty of scientific papers and wikipedia articles for any complex topic. The point of the article is instead to introduce the topic in plain english without extensive mathematical notation or expressions. The idea _is_ simple. Perhaps I should have added some visualizations of the Pareto front, but I think sometimes these graphs are shown unnecessarily quickly. Besides that what would you add to an introduction that is paramount to one’s understanding?

I wonder when the transition will happen and when DALLE 2 and similar technologies will actually replace jobs. Because practically, they can already.

One benefit with genetic algorithms is the fact that it can handle multiple objectives, like the NSGA-II algorithm. I used it to evolve a neural net in my master thesis.

I had an ad on facebook for a website and I saw a lot of attention on the post in form of likes and even comments but almost zero clicks to the actual website. I'm never buying another ad there, that's for sure.

While RL is about the problem it's also about the solution. Problems/Environments are formulated in a way where methods can be applied easily (i.e. Markov decision process) and thus the solutions are directly connected to the way the problem is formulated.

Deep learning is used for function approximation and is not in contrast with evolutionary computation. You can train a neutral network policy (mapping states to actions) with an evolutionary algorithm, but most of the success has come from methods that utilize the internal structure of the problem as mentioned earlier and evolutionary algorithms do not, which is what makes these optimization strategies both weak and powerful.

I've also made a StumbleUpon alternative called Viralwalk:

https://www.viralwalk.com

It's cool that many are attempting to recreate it, it's a shame it disappeared.

For online learning, multi-output, non-negative output, unlabeled data etc neural networks works well. The power of deep learning lies in how you can shape the problem and loss function for specific purposes. And even if these circumstances do not exit they can do well, it's all problem specific.

There's no rule that when the number of parameters is small deep learning shouldn't be used. The one time where deep learning maybe shouldn't be attempted at all is when the number of samples is very limited. While it excels with high dimensional hierarchical data it can do well on other problems as well. It differentiates from problem to problem and usually multiple solutions are tried and compared, starting with EDA and linear regression.

I guess you're right. Everything has a snowball effect on the internet, the hard part is getting the ball rolling.

I'm trying to understand what makes this app so popular, I created my own notetaking-app and did a shown HN and got like 5 upvotes yet I have plently of features this doesn't have like nested folders, search, hotkeys, code tagging, tools for creating tables/code, uploads of pdfs, images and all of this completely free. It's almost like setting a price on the service makes people want to use it more. </rant>

Thanks, have worked out so far.

To me the most important thing in notes is to find what you want easily. Because of this I think having some sort of online system is needed, otherwise if you have the notes at different places you will in the end give up on it and also tags and search is a huge advantage. But in the end I don't think there are one system that fits all, Zettelkasten might work for some people while others want it another way, this hypothesis is strengthened by the continuous output of new notes software. For this reason I created my own system, fitting _my_ needs. As a coder I chose markdown to also have the functionality to add snippets and search those snippets. The markdown documents stays in a folder type structure to organize the content and be able to use the vs code command cmd+p that I'm very used to. This works for me, https://www.theorylog.com.

I built something similar for my latest webapp to validate json-requests in python, the syntax is like this:

    {
        "content": Required(str),
        "username": Required(str, validate=username_is_ok, transform=lambda x: x.lower(), 
                   fail_message="Username isn't valid"),
        "message": Optional(str),
        "some_list": Required({
            "name": Required(str),
            "date": Required(str)
        }, is_list=True)
    }

Then I can provide this in a hook to the request method.

Hello everyone,

For a long time I thought there must be a better way to utilize notes in your work. Physical notes are difficult to bring with you and "search" through and ordinary note apps often don't have functionality for math and code snippets.

For this reason I created this website where you can add the following:

Links:

The key thing here is that these links are searchable by the content they point to since they are scraped. This is useful for things that are hard to google and you want to keep for later use.

Markdown documents with additional functionality:

Here you can add:

* code snippets that are made searchable, independently of the markdown document, by adding description and/or tags.

* images

* math (latex)

* PDFs (converted to images)

The text inside the images/PDFs are also extracted and added to improve the search of the markdown document (this doesn't happen instantaneously though but can take a bit of time to propagate). The markdown documents are useful when taking notes since you can combine theory and code snippets while making both searchable.

Still in the early stages, feedback is appreciated!

I'm currently working on a markdown system where I wanted to extend codeblocks to be able to incorporate more information. The `js:title=example.js` just doesn't cut it. Want I landed on was some sort of yaml-ish format:

    ```
    ***
    property1: ...
    property2: ...
    ***
    code
    ```

This way multiline properties are supported and I can also incorporate markdown inside the properties themselves.