azeirah

12 ms·

121.
by azeirah 2y ago
I sshed into my personal site vps this morning, created a file with vim, pasted in the html and added some styles by copying them from a php file in the same directory. All on the server! Do I get nerd creds now? <3
122.
by azeirah 2y ago
Yes-ish, I somewhat agree with your critique. I think it's a temporary measure though. It is very America-centric and that's very sad to me. I think Bret is a bit hesitant to share the stuff before people understand what it is, to
123.
by azeirah 2y ago
The books are the books that are recommended to read if you want to understand what they're doing and what prior work they're building on. Bret mentioned computer literacy on the website. The books you can consider as "the cu
124.
by azeirah 2y ago
Despite appearances, dynamicland isn't about any particular paradigm on its own. It's a set of values and a way of looking at media that come together to create computer-literate communities and even culture. This is obviously a b
125.
by azeirah 2y ago
Awesome! I'd love to see this flourish.
126.
by azeirah 2y ago
I like stickers I will contribute :o
127.
by azeirah 2y ago
:) I love this space of hybrid computing. If you want to see some inspiration, look at - bill buxton at Microsoft research - Disney research has a lot of really cool stuff too - anything siggraph Please keep me up to date one way or another
128.
by azeirah 2y ago
Logitech recently released a pen for this kind of purpose. It's meant to be used with virtual reality headsets, I'm not sure if their pen can be made to work outside VR. Do you need help with integrating your app into blender as a
129.
by azeirah 2y ago
From my perspective it is neither wrong nor right. Articles are supposed to be written for a target audience, and I think the only issue is that the original article didn't mention the target audience is people who don't want to a
130.
by azeirah 2y ago
This is really cool! I'd love to see more of these innovative fonts that justify their place in the world
131.
by azeirah 2y ago
100%, hacking is as much technical prowess as it is social engineering.
132.
by azeirah 2y ago
You can store information with basically any object. You can count in binary using your fingers ¯\_(ツ)_/¯ You could use pine needles to count. You could do steganography in knitting. You could encode information in the dust collecting
133.
by azeirah 2y ago
Yes. Tacit knowledge is hard to teach.
134.
by azeirah 2y ago
Jetbrains' recent WriterSide tool is a great addition into this space. It's not exactly wysiwyg, it's side-by-side live preview with different source languages (right now, markdown and some xml language are supported, but the
135.
by azeirah 2y ago
I'd recommend starting with the book "The nature of code", if I recall correctly, a new version was published very recently
136.
by azeirah 2y ago
I don't think that's what they mean. They explicitly stated that to achieve the current results, they had to manually translate the problem statements into formal mathematical statements: > First, the problems were manually tra
137.
by azeirah 2y ago
I'm not particularly experienced, but I did have good experiences with picovoice's services. It's a business specialised in programmatically available audio, tts, vad services etc. They have a VAD that is trained on a 10 seco
138.
by azeirah 2y ago
In practice, it's fine to stick with "just" 8k or 16k or 32k. If you're working with data of over 128k tokens I'd personally not recommend using an open model anyway unless you know what you're doing. The model
139.
by azeirah 2y ago
You don't need a 4090 at all. 16 bit requires about 24GB of VRAM, 8bit quants (99% same performance) requires only 12GB of VRAM. That's without the context window, so depending on how much context you want to use you'll need
140.
by azeirah 2y ago
I don't use an LLM in isolation. I use it alongside reading material, books, papers etc. Since I'm using it in involved studying sessions, it is totally sensible to cross-check most of what the LLM outputs, since that's part
141.
by azeirah 2y ago
I really hope to see more research like this. I feel like LLMs have insanely high pedagogic value, and that's because I've been using them to teach myself difficult subjects and review my understanding. The issue with the kind of
142.
by azeirah 2y ago
The conclusion you walked away with is the opposite of what usually works in practice. The more context you give the llm, the better. The key takeaway from that paper is to keep your instructions/questions/direction in the beginni
143.
by azeirah 2y ago
If this is novel and as reliable as I feel like they're claiming it to be, then this might be really big for svg-rendering in the web, absolutely amazing!
144.
by azeirah 2y ago
Have you tried Phi 3? It's smart which makes it perform well on benchmarks, but it's not great at conversation or as a chatbot. I imagine Gemma 2 is a better general-purpose assistant for most people, whereas Phi 3 is a solid smal
145.
by azeirah 2y ago
That's LLaMa-3-70B. The demo he gives at 6:09 is tinystories-15m, which is 30.4MB, so you'd only have to add the font to that (80~KB?) https://huggingface.co/nickypro/tinyllama-15M/tree/main
146.
by azeirah 2y ago
An idle game like cookie clicker can teach you much about what addiction could look like for you. It depends a lot on the person what kind of game will "click" and if they will "click" at all, but if they do you will be
147.
by azeirah 2y ago
You'll love this site :) https://treevis.net/
148.
by azeirah 2y ago
Huh, this is such an unusual design goal in programming languages "just try something and it'll probably work" Only css is done this way... but not even intentionally
149.
by azeirah 2y ago
I'm working on a tiny saas that is moving in this direction. It's based on laravel in the backend. I'm using different implementations of interfaces for self-hosted vs commercially (my variant) hosted. What implementations ge
150.
by azeirah 2y ago
I have some interesting ideas about this, as I've been thinking about it and have been prototyping ideas for ... a decade at this point? Do you have a way to contact you? Any reason you're not using treesitter with LSPs by the way

More ›