Code should be self-documenting. That way it is never outdated. Here's an example of this similar to what you can expect to see in practice:
def nabla_descent(X, y, theta, alpha, delta, nabla):
m = len(y)
for _ in range(delta):
h = X.dot(theta)
nabla = (1/m) * X.T.dot(h - y)
theta = theta - alpha * nabla
return theta
I (probably unreasonably) despise using web front-ends for desktop applications.
GTK is OK. QT is very feature rich, but that adds complexity. Both can be cross-compiled to most systems and shipped with all the required libraries pretty easily.
I haven't used it in a long while, but I remember liking Java Swing for some reason. Java should be "write once, run anywhere." But, cross-compiling isn't usually too hard, so not sure how much that matters. There's more modern frameworks for JVM-based languages now, but I haven't tried them.
I've noticed Gradio is popular in the ML community (web-tech based, and mostly used for quick demos/prototypes).
Edit: For web applications, I prefer Angular's more traditional architecture over React's hook architecture.
Texas Attorney General Ken Paxton is not letting up on efforts to stop the controversial Uplift Harris program.
Don't know why society tolerates these dumbass parasites.
Not sure I agree that there will be less human labor "need." Ideally, we should strive for progress, and not just survive. I think there is infinite use for human labor.
I agree with your second point.
I tried this with a faucet, but used agricultural 30% vinegar. It stripped the finish off, lol.
Edit: you can usually just unscrew the faucet cap, and replace the little aerator or soak the old one in vinegar. In my case, I was replacing the sink, so already had the faucet out, and there was lime build-up on the faucet itself.
IDK. Rocket Mortgage seems to be experts on being responsible with money, as evidenced by this company meeting: https://www.reddit.com/r/wallstreetbets/comments/1alzgv3/work_meeting_at_rocket_mortgage_time_for_puts_yet/
Haven't tried Gemini; may work. But, in my experience with other LLMs, even if text doesn't exceed the token limit, LLMs start making more mistakes and sometimes behave strangely more often as the size of context grows.
No problem. Glad you appreciated it. Namaste.
They indicated that they were wondering how pro-AI people would feel:
I really sat there wondering how do pro AI people feel when they get an email/SMS like that.
This is more complicated than some corporate infrastructures I've worked on, lol.
I usually just use VS Code to do full-text searches, and write down notes in a note taking app. That, and browse the documentation.
Nah, LLMs have severe context window limitations. It starts to get wackier after ~1000 LOC.
I'm pro-AI, but not pro-AI in the sense of, "turn these bullet-points into a verbose email," and not pro-AI for personal communication like this. I hope this kind of stuff doesn't become common. It's like going in the opposite direction of "SMS language" (which I view favorably).
Python is quite slow, so will use more CPU cycles than many other languages. If you're doing data-heavy stuff, it'll probably also use more RAM than, say C, where you can control types and memory layout of structs.
That being said, for services, I typically use FastAPI, because it's just so quick to develop stuff in Python. I don't do heavy stuff in Python; that's done by packages that wrap binaries complied from C, C++, Fortran, or CUDA. If I need tight-loops, I either entirely switch to a different language (Rust, lately), or I write a library and interact with it with ctypes.
With the Hispanic people I know that prefer Trump, it's the usual trumpist/Republican reasoning. Even down to anti-immigration, from a person who's father was an undocumented immigrant. Propaganda and desire to be in the in-group among your peers is wild.
"Fossil-fuel billionaire Kelcy Warren is about to land a knockout punch on Greenpeace..."
I don't think anyone is advocating for a "slap on the wrist." The U.S. criminal justice system is the most draconian in the West, and doesn't do "slaps on the wrist," unless you're in a particular economic or social classes.
IMO, ideally, he would be sentenced for as long as it takes to rehabilitate him. Could be 5 years, 10 years, 30 years, or never, IDK, I'm not a psychologist. But, the U.S. prison system isn't really designed for rehabilitation either.
Charging 14-year-old Colt Gray as an adult in Georgia school shooting reveals a nation that has forgotten the purpose of its juvenile justice system.
Production AI is highly tuned by training data selection and human feedback. Every model has its own style that many people helped tune. In the open model world there are thousands of different models targeting various styles. Waifu Diffusion and GPT-4chan, for example.
I think you have your janitor example backwards. Spending my time revolutionizing energy productions sounds much more enjoyable than sweeping floors. Same with designing an effective floor sweeping robot.
AI are people, my friend. /s
But, really, I think people should be able to run algorithms on whatever data they want. It's whether the output is sufficiently different or "transformative" that matters (and other laws like using people's likeness). Otherwise, I think the laws will get complex and nonsensical once you start adding special cases for "AI." And I'd bet if new laws are written, they'd be written by lobbiests to further erode the threat of competition (from free software, for instance).
It's going to take more than CAPTCHA to prove you're real
AI firms propose 'personhood credentials' to combat online deception, offering a cryptographically authenticated way to verify real people without sacrificing privacy—though critics warn it may empower governments to control who speaks online.
I use Google Shopping (the “Shopping” tab on Google) to see if local stores carry certain products, what they cost, how far away each store is, etc. It seems to mostly search national or large regional chains, but it was still pretty useful.
Is there any alternative to this (in the US)? The “nearby” function has unfortunately got shittier and shittier over the past year or so. It's gotten less “deterministic," just mixing results from local stores with e-commerce stores, further reducing usefulness.
I don’t remember how I heard of it, but just binged-watched it over the past few days. Ratings seem a little bit above average, but I found it very enjoyable. I liked that the mood oscillates between modern comedy and tragic comedy; and that it seems to implicitely critique modern society. The series almost feels like an allegory (or perhaps I’m reading too much in to it).
EliseAI, a startup developing AI-powered tools for property managers, has raised $75 million in a funding round valuing the company a $1 billion.
I've recently noticed this opinion seems unpopular, at least on Lemmy.
There is nothing wrong with downloading public data and doing statistical analysis on it, which is pretty much what these ML models do. They are not redistributing other peoples' works (well, sometimes they do, unintentionally, and safeguards to prevent this are usually built-in). The training data is generally much, much larger than the model sizes, so it is generally not possible for the models to reconstruct random specific works. They are not creating derivative works, in the legal sense, because they do not copy and modify the original works; they generate "new" content based on probabilities.
My opinion on the subject is pretty much in agreement with this document from the EFF: https://www.eff.org/document/eff-two-pager-ai
I understand the hate for companies using data you would reasonably expect would be private. I understand hate for purposely over-fitting the model on data to reproduce people's "likeness." I understand the hate for AI generated shit (because it is shit). I really don't understand where all this hate for using public data for building a "statistical" model to "learn" general patterns is coming from.
I can also understand the anxiety people may feel, if they believe all the AI hype, that it will eliminate jobs. I don't think AI is going to be able to directly replace people any time soon. It will probably improve productivity (with stuff like background-removers, better autocomplete, etc), which might eliminate some jobs, but that's really just a problem with capitalism, and productivity increases are generally considered good.
The Georgia State Election Board creates rules for the battleground state's elections, and its Trump-approved majority is trying to make changes.
O’Brien acknowledges Biden has been a “great” president for organized labor. But he told the Globe that Biden hasn’t delivered on all his promises and the Teamsters are worried their backing is being taken for granted.
An updated fable for 2017
As the energy transition inches through the ‘issue attention’ cycle, a wiser approach should emerge.
Proponents of the Greater Idaho movement have argued Democrats in Portland don’t understand their way of life
Sometimes experts get it wrong, but not this time.
Meta appears poised to spend billions of dollars by the end of this year on Nvidia's popular computer chips that are widely used for AI projects
Summary: Meta, led by CEO Mark Zuckerberg, is investing billions in Nvidia's H100 graphics cards to build a massive compute infrastructure for AI research and projects. By end of 2024, Meta aims to have 350,000 of these GPUs, with total expenditures potentially reaching $9 billion. This move is part of Meta's focus on developing artificial general intelligence (AGI), competing with firms like OpenAI and Google's DeepMind. The company's AI and computing investments are a key part of its 2024 budget, emphasizing AI as their largest investment area.
I'm out of room for breakers in my main breaker box, so would like to add 6-breaker sub-panel to install mini-splits, an outdoor electrical outlet, lighting, etc. What's the correct way to mount an exterior sub-panel on a house with lapped hardie-board siding?
I suppose the easiest thing would be to drill holes in the siding then attach the panel with screws to a stud and the exterior sheathing, but I don't know if that's a proper way to do things.
I also suppose I could somehow cut a rectangular hole in the siding and mount the subpanel directly on the sheathing. I'm not sure how to prevent water intrusion in that case (is some kind of flashing needed, or is just cock ok?). Seems like it would be hard to cut a clean rectangle in lapped fiber-cement siding on a vertical surface.
I'm working on a prototype application, and for my use-case DeepFloyd IF gives me the best results by far. I was using replicate.ai, but the reliability and cold start times are unacceptable for my purposes (and DeepFloyd IF cannot be used in commercial products).
I think what made IF so good for my use-case is that it starts by generating a very small image. I need images generated with a single subject taking up most of the image and minimal background. I think IF is biased to these kinds of images since it starts with a very small image.
Currently using Dall-E 2, which is ok, but not nearly as good as IF. Stable Diffusion sometimes produces very weird images (haven't tried SD XL yet).
I'm seeing strange behavior when I click on a post, then click the "back button" in my browser. Sometimes if I'm on the "subscribed" tab, click on an article, then press back, it seems to show me "all" or "local" posts. Sometimes it shows me a different list if I'm on the "all" tab, click on a post, then press back. Same behavior on Firefox mobile and desktop version.
Haven't went into in-depth testing, but I can't be the only one seeing this right?
Guessing it's something to do with browser, CDN, or server-side cache?