simbolit

This person has no problem with "AI", this person has a problem with capitalism.

Because they formed an additional entity, which is a for-profit, and has leadership treating the whole thing as a for-profit, and so employees also see what their eyes are telling them.

The misalignment is not accidental, it was carefully developed over the last few years.

That's fun.

Change the name while you are at it; the company is not any more "open" than the next shop.

You need to provide an email address and click on a link. Emails from grr.la work.

Yes, it is "register an account", but it is the lowest friction I know.

me reading your comment: yeah, also hated that, yes, yes, yes

your last four words: nope, hard no, no, we are not friends

> Good execution remains differentiated. It just requires continuous iteration, evolution and improvement.

But with the underlying basis for everything being open source, everything you learned with "continuous iteration, evolution and improvement" can be copied relatively easily.

TLDR: The underlying basis is the same for everyone:

# the whole internet to scrape

# the largest amount of gpu compute you have ever seen

# more or less open source fundamentals

thus "ai" will become a commodity, unless you have specific non-public useful data.

That's the first section of the article. It then continues for four more sections.

FYI: this is not 'fusion power'

"first-of-its-kind Multi-State Fusion (MSF) Reactor which can be used to create vital medical isotopes. These isotopes can be used for killing cancer cells in radiotherapy and as radioactive tracers in the body to help spot medical conditions through imaging."

If you don't add an antenna to your airgapped computer, you need physical access to some copper.

If you do add an antenna to your 'airgapped' computer, well, just don't.

Is there a third way?

IIRC from about 2005-2010 there were intense experiments as to using torrents for video streaming. But, again iirc, the results were that there was too much competition for the earlier blocks. This was because unlike conventional file sharing, streaming is linear and people abandon videos halfway. Ultimately the savings over conventional hosted streaming were small. And so Blizzard and Nine Inch Nails used it, but the video players didn't.

Do I misremember something, or has something substantially changed since then?

Interesting to see the "try for free" tier using gpt3.5t, and the paid tier using gpt4t.

"Here you can try something very much like our product, but our product is actually better" is not the most orthodox marketing method. I'd say giving a small sample at full quality is very standard. Is this an ML thing?

FYI: This is a proof of concept, not a product, not a product announcement.

Doesn't "being different and ok with it" include some people wanting to do something "really stupid and dangerous" and you being ok with it?

Could you please expand on your "if you think about it"?

think about video compression for a "zoom" call.

You could just send a lot of bitmap files. Or, you could save bandwidth and zip them before send.

Or you could integrate some image-specific compression into the file format, as in motion-jpeg. And in mpeg2/h264/h265 you supercharge this with temporal compression, not just adjacent pixels, not just blocks within the frame, but also pixels and blocks from adjacent frames are used to predict each pixel.

And now think about not sending video, just a single initial jpeg, some face shape data, and now a continuous stream of position data of a dozen or so points on your face and the facial movements are reconstructed. (Nvidia research project from last year)

And now think about no longer sending images at all, just coordinates for points in latent space (aka very abstract image descriptions) and a super-fast diffusion model "generates" the right frames on the fly.

Where does "compression" end, where does "AI" start? Ultimately, the marketing people decide.

Don't get hung up on the term "AI". "AI" is a bullshit hype buzzword. Calling it such serves no scientific-practical purpose, it is solely meant to impress venture capital.

they significantly changed the "darkplates". The 2.0 lawyer-approved version doesn't even have the distinctive "fins" (or collar or whatever you want to call it) .

Dbrand used trap streets :)

https://en.wikipedia.org/wiki/Trap_street

user @bArray 35minutes ago:

"Tested on a NVIDIA GeForce RTX 3050 under Ubuntu with 4GB VRAM. (...) lowered the canvas to 2Kx2K and it seems to just about be okay. My test prompt (...) produces a picture of rocks. (...) I get a nice scene (...) Both take about two minutes."

If you look at the very right of the screenshots, there is a "history" of generations with unused alternatives. Me and my visual cortex managed to synthesize "before" images from this information.

Very inconvenient? Yes. Completely useless? No.

>find your nearest art college and go to the end of year show

This. I just want to add: If you want to see more interesting things, if you have a choice, go to the less prestigious school. I have seen my fair bit of end of year shows, and the prestige of the institution is directly correlated with how orthodox everything is.

Why are journalists so often so terribly ill-informed? You'd think it's their job to be well-informed, but apparently not.

>require months of re-training

the computers they have, you wouldn't believe it...

I am no longer sure there is much disagreement... just a different perspective on the world, I guess?

What you say is kinda true, but directly proportional to time since the last commit (currently for uBlock 9hours ago). If there has been a commit over the last few days, yes, I (not particularly reasonably, but still) expect there to be updates. But this expectation (and reasonableness) diminishes with this commit further away, if that becomes months and then years, I no longer expect that.

And in your hypothetical conversation, a lot hinges on your exact framing and phrasing of both the (non)entitled question and the phrasing of the answer. Yeah, your hypothetical developer is a dick. Don't be a dick, both sides.

Again, I feel there isn't enough disagreement here for argument. Enjoy your perspective, have a good day or night.

Let's look at one of my favorite open source projects:

https://github.com/gorhill/uBlock

Does gorhill need a disclaimer, as you propose?

If so, please tell me specifically, where you are reading any "implied support, security bugfixes, and future development".

I am not purposefully complicated, I just feel your "implied" is doing a lot of work, and I want to see what that really means. I don't see the implication, but perhaps I am blind. So please enlighten me. (If ublock doesn't fit your argument, please give another example of your liking.).

I am not sure what you are saying here.

Is this is a general critique of capitalist exploitation of the "software" field? Like, as in software should be a utility like fresh water or electricity?

If not, and you are relatively literal in what you write, the following question seems unavoidable:

What about coaches for youth sports? What about beach cleanup? What about a thousand other worthwhile and societally useful activities that people volunteer to do, but for which there is no sustainable career path? What makes OSS contributing any different from them?

You are saying I should add, to my free lemonade stand, a disclaimer:

I will not help you with drinking the free lemonade; if the lemonade is too sour for you I will not be providing extra sugar; and I will not make free lemonade tomorrow.

Can you give me one or two examples of official OSS websites where these are heavily implied or outright stated? Do you mean anything beyond a roadmap?

That sounds Gibsonesque.

this is meant to hold back regulators, just for a couple more years, so, for a couple more years it will be green bubbles android, blue bubbles iOS.

The teenage market share is 87% currently. just a couple more years is all they need.