itissid

11 ms·

121.
by itissid 1y ago
The book "Apple in China" by Patrick McGee focuses on this point in early 2000 top 5 contract manufacturers were in north america. By around 2010 the top one was foxconn which was larger than the next 4 combined. This is going to
122.
by itissid 1y ago
My experience with Claude code beyond building anything bigger than a webpage, a small API, a tutorial on CSS etc has been pretty bad. I think context length is a manageable problem, but not the main one. I used it to write a 50K LoC python
123.
by itissid 1y ago
Human goals are more important. I think conceptually the idea should always be strong goals set by humans and then sub goals each with a praticularly well defined *plan* for meeting them. This needs to be the conceptual basis, if you are ha
124.
by itissid 1y ago
A few things need to happen very soon(if the signs are not here already): 1. Tech Company's should be able to accelerate and supplant the FAANGs of this world. Like even if 10x was discounted to 5x. It would mean that 10 human years of
125.
by itissid 1y ago
I think an accident mixing of 2 different pieces of info, each alone not enough to produce harmful behavior, but combined raise the risk more than the sum of their parts, is a real problem.
126.
by itissid 1y ago
Who is doing all the work of making physical agents that can behave as good as a UBI generator? Something that can not just create videos, but go get groceries(hell grow my food), help a construction worker lay down tiling, help a nurse fe
127.
by itissid 1y ago
IIUC, we need gibbs sampling(to compute the weight updates) instead of using the gradient based forward and backward passes with today's NNetworks that we are used to. Any one understand why that is so?
128.
by itissid 1y ago
There is a giant indoor climate controlled greenhouse which allows visitors that has a heated patio in the middle to sit. Makes a great place to lounge in the winters.
129.
by itissid 1y ago
When you frame it as an optimization problem, like by optimizing the squares loss or cross entropy, you have decided that your data generating process(DGP), i.e. Y is: - A Binomial/Multinomial random variable, which gives you the the c
130.
by itissid 1y ago
Another way to approach the explanation is understanding the data generating process i.e. the statistical assumptions of the process that generates the data. That can go a long way to understanding _analytically_ if linear regression model
131.
by itissid 1y ago
I've been using it for pre scoping things I have no idea about and rapidly iterating by refeeding it a version with guard rails and conditions from previous chats. Like I wanted to scope how to build a home made TrueNAS Scale unit, it
132.
by itissid 1y ago
Permissioning is a mess – among other things of course. I feel that permissions to access any resource(image, location, files etc) could be given without necessarily giving access to the PII value that resource holds, e.g. running a no side
133.
by itissid 1y ago
Isn't this kind of another way of how Inference Time Scaling works? It will basically produce several chain of thoughts and then pursue one that has maximum reward based on an internal function?
134.
by itissid 1y ago
> and I want a 3mm hole in one side of the plate. No the other side. No, not like that, at the bottom. Now make it 10mm from the other hole. No the other hole. No, up not sideways. One thing that is interesting here is you can read faste
135.
by itissid 1y ago
This can only be true of some products. Often there are a lot of concerns like privacy, white labeling, legal consequences that need to be considered _before_ you vibe code.
136.
by itissid 1y ago
Not all data. Not storing location history data is an example of not opening this for gestapo by ommision. For e.g. Apple does not furnish user location info on geo fence warrants because it can't. I believe Google just made a change t
137.
by itissid 1y ago
For home improvement projects, This could be quite useful for generating point cloud map of places hard to get to. Like I have drywall installations I would love to get behind and check how things look, this would be great for that.
138.
by itissid 1y ago
Why does yolo not work?
139.
by itissid 1y ago
You know if we could this open source self hosted stuff with a layer of Patreon on top people would pay something to all the people. If one could make it non-intrusive it could become a decent alternative to paying all these artists. I thin
140.
by itissid 1y ago
Why is Ultralytics yolo famous for it?
141.
by itissid 2y ago
I've been running an ollama and deepseek in a container in TrueNAS k8s for several months. It's hooked up to my Continue extension in VSCode. I also mix it with cloud hosted "dumb" ones for other tasks like code completi
142.
by itissid 2y ago
Nice!. I would like to do something like this but for Photos. I want to get off Dropbox and Google for a while. At the top of my list are: 1. Create a memories like feature from my photos to send on my cell phone, grouping by different feat
143.
by itissid 2y ago
My idea was to convey was that the sense of identity attached to a people is weird. And that people are just people. You can get the same set of people(genetically, demographically same) to kill millions in a war and become pious afterwards
144.
by itissid 2y ago
Its just people. People are the same everywhere, and are fundamentally unpredictable systems. How large groups behave does depends to a certain extent on context: by compared to others and your socio-economic situation. How they publicly ex
145.
by itissid 2y ago
I can relate to this pain. Isn't gitlab CI better at this especially the documentation and simplicity of it?
146.
by itissid 2y ago
Travel time context in general could be useful for retrieval before ranking in searches like Yelp or Google maps like products for nearby events and places. My MVP use case is to configure a rerank based on commute "score". For e
147.
by itissid 2y ago
Since I already have a browser connected to the Internet where this would execute, could one have the option of transparently executing the webGPU + LLM in a cloud container communicating with the browser process?
148.
If C++ cmpler became smarter would't be opinionated as rust borrowchecker?
4 points by itissid 2y ago|2 comments
149.
by itissid 2y ago
Woodworking projects are the hardest of these. I've installed my own under the sink RO and it's relatively self contained 2 hour work whenever I move. Yet there are unforseen like tighter spaces, old plumbing that needs replacemen
150.
by itissid 2y ago
I call this " why the cost of hiring a guy to do anything is either 250$ or 10000$ ". And when the switch from low to high happens your jaw drops. Most routine maintenance work is the latter coat point(your toilet flus repair, ele

More ›