la_fayette

ok cool, i see there is already a similar project called wanderer, so i assume the difference would be in the planning module? thx

I started to work on an open source audioguide app as PWA or mobile app. This app can be used by museums or similar institutions https://github.com/smartcompanion-app/audioguide-app.

After my experience in consulting museums, I found that they spend a lot of money on software, which could potentially be replaced with open source software. Therefore I started an awesome list with FOSS software for museums: https://github.com/smartcompanion-app/awesome-open-source-mu....

i am not sure acctually of the math is acctually that complicated/important. the math around neural networks is calculus/chain rule etc and for model comparison/validation one needs statistics. the required math for e.g. understand transformers is quite accessible.

When reading your comment, it just reminds me on how feature flags can be misused as application configuration/customization. An antipattern i could observe at various organzations already.

For me feature flags go along with trunk based development to enable features in QA settings, but not on PROD yet, for PO/PM testing. Trunk based development allows for fast/easy devops, without complicated branching strategies.

Application configuration is, for me, part of the application and has the business context for customizing the application accordingly. Not sure if there are specific frameworks/tools out there. But one should clearly distinguish these two.

depends, JavaScript in the Browser has many useful things available, which I miss with python, e.g., fetch, which in Python you need a separate package like requests to avoid a clunky API. Java had this issue for long time as well, since Java 11 there is the HttpClient with a convenient API.

Yes, that’s true! That could be my next step… though I have to admit, writing this in a HN comment feels like a bit of a challenge.

This guy is so amazing! With his video and the code base I really have the feeling I understand gradient descent, back propagation, chain rule etc. Reading math only just confuses me, together with the code it makes it so clear! It feels like a lifetime achievement for me :-)

The claim in the article that AI is good at writing code, is that claude code is written by claude code.

I don't get your point here. User targeted ads are the main business model of the internet? Yes, few days ago it was revealed how billions of user data points could be gathered from Meta [1], did anybody care, outside a small privacy community? So indead these things are not surprising... My thoughts don't go so far to consider the effects on society, idk, do you?

[1] https://www.heise.de/en/news/3-5-Billion-Accounts-Complete-W...

It is no surprise, somehow they need to earn money. It will be interesting though how much the response of the LLM will be adapted. At least legally advertisement need to be marked for users. So either the response of an LLM will be extended with ad content or replaced by ad content.

I use GenAI for text translation, text 2 voice and voice 2 text, there it is extremely useful. For coding I often have the feeling it is useless, but also sometimes it is useful, like most tools...

https://github.com/smartcompanion-app/audioguide-app

An open-source audioguide app that helps museums and cultural spaces create engaging visitor experiences. Feel free to give me a star on GitHub.

It might not be possible but others have scraped just the app store and matched based on Android meta data: https://people.ece.ubc.ca/amesbah/resources/papers/mobilesof...

Somebody scraped the play store and checked the framework, so a list for Android WebView apps, built with capacitor, is here: https://capgo.app/top_capacitor_app/ Maybe an equivalent is there on iOS for the same app...

Let's see what will the future of Graphene be, since Google is not publishing the device tree anymore for Pixel devices...

Thinking without idiology about it, it seems to make sense to have an autonomous agent with a wallet. A requestor could send money (in whatever form) into the wallet of the agent and would buy tokens, processing time or whatever from the agent.

I can recommend Apache Camel (https://camel.apache.org) for similar data integration pipelines and even agentic workflows. There are even visual editors for Camel today, which IMHO make it extremely user friendly to build any kind of pipeline quickly.

Apache Karavan: https://karavan.space/ Kaoto (Red Hat): https://kaoto.io

Both are end 2 end usable within vscode.

Although I didn't collect numbers, but I made a similar experience in my workplace. I assume many people are highly distracted by ads and work efficiency is even reduced. Even many software engineers seem to not be aware of ublock... Would be interesting to know how many students started using an ad blocker at the end of your lecture :)

Great, I will check it out, I am a Java dev and always wanted to learn about game programming!

I am working on an open source audioguide app for museums and similar institutions https://www.smartcompanion.app/

Feel free to give my repos a star on GitHub, thx

Since such a database should evolve continuously, I wouldn't see that as a problem. The important thing is, that each example is somehow verifiable, in the form of a unmodifiable test setup. So the LLM provides a solution, which is executed against the test to verify. Something like ACID3 Tests... But sure it can be gamed somehow in probably all setups...

At this point it would be an interesting idea, to collect examples, in a form of a community database, were LLMs miserably fail. I have examples myself...

I think HN might be one of the few communities where people have been running extensive experiments with LLMs since their inception. Most here take a realistic view of their capabilities. There are certainly proven use cases where LLMs provide clear productivity gains—for example, copying an error message and retrieving potential solutions. At the same time, many recognize that marketing fantasies, such as the idea of having a "PhD in your pocket," are far beyond what this technology can deliver.

Yes that sound like important and useful use cases. However, these are solved by boring old school ML models since years...

Does anyone know how Norvig thinks about the use of LLMs for programming?

In general I would agree, however the resulting systems of such an approach tend to be "just" expensive workflow systems, which could be done with old tech as well... Where is the real need for anything LLM here?

When you argued that it saved you time by a factor of 10, have you even measured that properly? I initially also had the feeling that LLMs save me time, but in the end it didn't. I roughly compared my performance to past performance by the amount of stories done and LLMs made me slower even if I thought I am saving time...

From several month of deep work with LLMs I think they are amazing pattern matchers, but not problem solvers. They suggest a solution pattern based on their trained weights. This even could result in real solutions, e.g., when programming Tetris or so, but not when working on somewhat unique problems...

What exactly does bitcoin offer the world today?

It is a highly reliable, global-scale P2P software system, we can analyse, experiment with and learn from.

Ok but the association of BYD with China is also something I personally have reservations with.

Unbelievable that this book is freely available! Thanks to the authors, publishers or whoever.