loudmax

31 ms·

361.
by loudmax 3y ago
I presume most of the concern with "offensive" content is CYA from Anthropic. You or I may not be offended if an AI generates stupid content, but some people would be, and some people will go out of their way to be offended. Ant
362.
by loudmax 3y ago
Yes. There's no getting around the fact that the sun rises and sets around the world at different times. What we do instead is get rid of the assumption that everybody works 9am to 5pm local time. It does require a different way of t
363.
by loudmax 3y ago
Fully agree that the existing copyright and intellectual property systems are dire need of deep reform. But to get people on board, you can't just propose burning it all down, you need to point to a viable alternative. Say, limiting
364.
by loudmax 3y ago
The two-stage model sounds a lot like the "Hires fix" checkbox in the Automatic1111 interface. If you enable that, it will generate an image, and then generate a scaled up image based on the first image. You can do the same thin
365.
by loudmax 3y ago
The Vision Pro looks like a lot of fun but I'm not sure what the use case is. Unlike the iPhone, this is clearly an indoor device, not an AR system to navigate around town. It seems a much more immersive desktop experience, but are you
366.
by loudmax 3y ago
It might depend on what you mean by "full training" and "fine tuning". They're not proposing to train a brand new foundational model from scratch, like a brand new LLaMa. But they want to do something considerably
367.
by loudmax 3y ago
Also looking forward to Stable Diffusion's new SDXL model in "mid-July": https://stability.ai/blog/sdxl-09-stable-diffusion I hope both of these deliver on their promises. They're exciting developm
368.
by loudmax 3y ago
I recommend the Huggingface courses: https://huggingface.co/course The courses are geared toward writing Python applications around these models. They're fairly hands on, so it would still be a good thing to complemen
369.
by loudmax 3y ago
It might be technically possible to port CUDA to Intel or AMD, but there might be patents of copyright rules that prevent legal redistribution. This wouldn't be the first time that IP rules stifle free market competition. Probably bet
370.
by loudmax 3y ago
Are tracking memory allocation or variable types not pain points for large complicated programs?
371.
by loudmax 3y ago
I think the hope is that Red Hat would care about its own long term profitability. That is, the people running Red Hat will realize that they're sacrificing future revenue by prioritizing instant profits and this course is unsustainab
372.
by loudmax 3y ago
Check out the Huggingface NLP course: https://huggingface.co/learn/nlp-course/chapter1/1 Huggingface has a bunch of courses, but that's a good one to start with. You can do the exercises on your own com
373.
by loudmax 3y ago
AI development is taking place at breathtaking velocity and GPT-4 isn't the end of the road. OpenAI's moat relies on them improving at a faster pace than competitors. I wouldn't take it for granted that OpenAI's lead w
374.
by loudmax 3y ago
From their project page https://github.com/openlm-research/open_llama : > For current version of OpenLLaMA models, our tokenizer is trained to merge multiple empty spaces into one before tokenization, similar to T5
375.
by loudmax 3y ago
As someone who admired the old Google, I totally agree. It's terribly disappointing to see so much potential go to waste under the current leadership. OpenAI was supposed to be a hedge against Google's monopoly of artificial inte
376.
by loudmax 3y ago
If that's the case, they've overshot the mark. Sometimes it's better to lose money on a failed product to reassure potential customers that you can be relied on as a long term business partner.
377.
by loudmax 3y ago
I think that is a reasonably analogy, but an important difference is that a startup has a strong incentive to be committed to their product. A startup may not stay in business for long, but we have an expectation that they'll support
378.
by loudmax 3y ago
Yes, and look at the extremism and social delusions and social networking addictions that have been exacerbated by the internet. On balance, it's still positive that the internet exists and people have open access to communication. We
379.
by loudmax 3y ago
I tend to be sympathetic to arguments in favor of openly accessible AI, but we shouldn't dismiss concerns about unaligned AI as frivolous. Widespread unfiltered accessibility to "unaligned" AI means that suicidal sociopaths
380.
by loudmax 3y ago
I suspect this isn't the direction that System76 is pursuing with Virgo, but Linux is in a good position to take advantage of Arm CPUs. Linux isn't tied to legacy software that only runs on x86 in the same way that proprietary OS
381.
by loudmax 3y ago
Developments in this space are coming really fast, and reading words are squarely within the capabilities of neural engines. 5 years is a very long time in AI years.
382.
by loudmax 3y ago
My plug-in Prius gets about the same range. If that range is sufficient for most of your driving, I think PHEV is a fantastic compromise between driving electric most of the time and switching to ICE when you need the additional range.
383.
by loudmax 3y ago
> The universal language and portability of a spreadsheet Depends strongly on your definitions of "universal" and "portable". Try reading these files with non-proprietary software.
384.
by loudmax 3y ago
YouTuber Perun, who does videos on defense procurement, has an episode on Hypersonic missiles: https://www.youtube.com/watch?v=0n3fjoacL20 It addresses a lot of the same topics as this Medium article.
385.
by loudmax 3y ago
I wonder about the copyright status of code put out by ChatGPT. I presume anything ChatGPT generates belongs to the user who asked for the code to be generated. Is there any clause in the EULA that says this is the case? Or is code genera
386.
by loudmax 3y ago
I think the 1-2 years estimate is a bit short. Right now, a computer can create a droning audiobook out of a written book. Within the next few years, generative models will create cheap mediocre audiobooks. But the technology is changing
387.
by loudmax 3y ago
I'd argue that a file manager shouldn't be a function of the Desktop Environment. No that a GUI file manager is necessarily bloat, but a file manager should be independent of whatever DE you're using. Tiling window managers
388.
by loudmax 3y ago
Here are two: * Elaine Luria * Barbara Comstock Both are from Virginia, and both were voted out of their politically moderate districts. This is the tragedy of being a centrist. Most Republicans aren't deeply racist and most
389.
by loudmax 3y ago
Most American voters are moderates. Party primaries and gerrymandering produce politicians that reflect the most active elements of the party base, rather than the majority of party voters. Yes, you can still find moderate politicians if
390.
by loudmax 3y ago
I disagree. Most Republicans and Democrats are moderates, and the polarization is coming from loud vocal minorities at the fringes. Republicans and Democrats erroneously assume that the other party's median is far more extreme than th

More ›