davidy123

Hi, I've helped several "startups" get off the ground, because I'm good at creating rapid prototypes that scale to production, and I like solving big problems, with contributions based on informed imagination. I've experimented a lot with AI, the practical type that doesn't require deep re$earch. I'd like to find a funded trial balloon type startup, where I can be a hands-on technologist for ideation, full stack software. I'm particularly interested in projects that succeed by doing what closed companies can't. Currently ET but soon to be PT, you can reach me at vid_hn@zooid.org. Thanks for your consideration!

«In one famous example of the flaws of fMRIs [Harris' PHd topic], researchers used a dead salmon as their fMRI subject. The salmon was shown a series of images of various human social situations, designed to evoke an emotional response. The researchers found that, using the standard methods employed by neuroscientists and psychologists, the dead salmon responded to the images, illustrating the insanely high false-positive rate of fMRI research.

[The statistician William Briggs, having studied the thesis] summarises:

“During the course of my investigation of scientism and bad science, I have read a great many bad, poorly reasoned papers. This one might not be the worst, but it deserves a prize for mangling the largest number of things simultaneously.”»

They claim PCIe4.0 makes it irrelevant, but that doesn't really make sense and it's most likely the case they want to charge a fortune for their high-memory options.

I mean semantic web is designed to turn the entire web into a semantic database. You could treat "your" parts of the web as DSLs in the simplest possible RDF/linked data and then link it to semantic web concepts (and other people can link it as appropriate), but I don't think you can create anything like a semantic web with standalone DSLs.

> That is what I was saying. Layouting and rendering general case graphs is a mess.

Yes, unfortunately and somewhat ironically since the web is a non-tree graph.

Your approach doesn't apply to the semantic web. Apples and oranges. Two dimensional non-tree graph are neat but in most cases at best a debugging tool or whirlygig.

Hmm I wonder if that's also because many more people are involved, which is a good thing. After all, the Web started as a research system, the more it returns to that the better. Intuitively I'd think LLMs/other algos could help since they can generalize, extract uniqueness, identify power laws, though that will be possibly gamed, that also helps create more robust algorithms. And it has a fascinating relevance to all works of society.

I guess for anyone interested in Bun and Javascript, it would be more obvious.

User what does "TCO" mean in " Bun, JavaScript, and TCO"

ChatGPT "TCO" in "Bun, JavaScript, and TCO" refers to "Tail Call Optimization."

I don't think it will be rendered valueless, but it shouldn't be completely exclusive. Basically as an extension of today's search model (which is a large part of what LLMs are, along with a grab-bag of useful ML algorithms), people should be able to access information universally, but if they want to go to perspectives or very fine details, then a pay model is acceptable, as long as there's a trail from freely available information and evaluable models. Ultimately imo there are bigger problems with elite/gatekeepered information than finding new ways to produce or support information development, given the power a few corporations are gaining and the opportunity to overcome stratified society.

That's not as true as it used to be, and there are still plenty of useful open journals/open science publications, though proper attribution would often be important.

[edit] you could pretty much say that on principle, any significant development should have a publication in the open.

The solution could be great. I really don't like the way culture always goes to the same tropes, calling any potential innovation "out of Star Trek" (with attendant distorted expectations), right down to expecting an interface based on literal hand-waving in Minority Report. If copyright held works ("USS Enterprise") could be removed, yet the actual essential concepts (space ship, naming things) retained, it would be a tremendous breakthrough.

I think what NYT &c want is for large companies like Apple to pay them for access to their works. This to me is the wrong path, just leading to more silos and walled gardens, special access for the elite.

An alternative is base models trained on Wikipedia and public domain (science journals, etc). Foundations could support high quality, well rounded current events reporting. Wikimedia provides a good model for this, with referenced summaries that I don't think can be said to reasonably violate copyright. The models would need to be improved to support references, or RAG attribution would have to be widely used when bringing in works that have a current copyright.

Sure, https://i.imgur.com/drvwqhj.png

The displays are connected to a workstation. I use barrier between the systems (except the right hand locked down one).

Like anything else it's a constant work in progress, but I'm fairly happy with it.

You could set up screen sharing for any system that supports it, and on another system receive the stream and use an LLM(s) to run object ("cat, piano") and text detection on images and sound. Mac silicon can do that, but it's not as efficient or cost effective as a dedicated GPU. This approach would be system neutral since models work on (increasingly) any GPU including Apple's. On your workstation, you'd just have the overhead (and concerns) of the video stream.

I haven't seen this config mentioned yet, but I have two displays; a LG SDQHD 28" 8:9 2560x2880 on the left, and a LG 49" 21:9 55120x2160 ultrawide curved on the right, with a slight angle joining them in a "hinge."

I use multiple virtual desktops, for dev I have a terminal on the left and the IDE on the right. This was the best config I could find where I had enough on the screens I didn't need to flip between panes frequently, yet it wasn't overwhelming and lets the actual window view behind the displays through, and I can comfortably focus on key areas and look around with craning my neck too much.

If the displays were backing a wall, I might try an 8k 65" display, but I would mainly use the comfortable part of it, and let the rest be ambient info. That could be magnificent with an OLED display. Though I think it would make sense to be curved and they aren't common anymore. Who knows, maybe a VR headset will be a better choice in a few years.

I've been working a lot on the idea of ambient info. I use KDE's html wallpaper feature to provide a dynamic graph view of my environment.

Oh I see. I thought they were giving general advice but that makes sense. The person could have used ChatGPT to write it for them, if language was an issue or they were in a really bad mood after the experience. I do that sometimes.

Remember to stay … just kidding.

sorry, what am I missing? What is insidious about the last paragraph?

This is going to be a controversial position, but nobody is really arguing the other side, the benefits of ads. And if they were more accepted and integrated (a "perfect" federated model), maybe they could be acceptable rather than abrasive and irrelevant. Who is talking, with a measured perspective, about the shift to participatory, user produced content (yes, I know, a lot of it isn't good, but some is and it is available to many more people without barriers), and the benefits that brings? I just see hand waving about alternatives, some "good old days" ideas about media, and a lot of people accessing essentially free ad-sponsored content with generally minor inconvenience in a murky going-nowhere conflict.

I think as much as anyone Google brings alternatives; I pay for Youtube Premium and never have any ads on Youtube/Youtube Music (places where it would be particularly unwelcome), I get paid for Google Surveys, and I think with some work their federated model could be perfectly reasonable.

If you disagree, please say why, rather than simply downvoting.

Yes, but everyone else is vulnerable to not being able to run extensions. As mentioned, I just use it for news & general web site surfing, no accounts.

I use Kiwi, I take the risk for the ability to run my own extensions (though I use a two-fisted approach where I use Chrome for deep accounts). It's a shame it's not updated more often, it's an open source project I would support.

oh yeah, and this esbuild keepNames nonsense has lost me some hair.

Maybe I spoke too soon, cyclical imports problem seems pretty big.

Just wanted to say thanks for this recommendation. I'm not finished yet, but it seems like a huge improvement over all the wtf of jest, along with a lot of new niceties.

I don't see why it wouldn't work, but the right thing to do is use accessibility features.

This doesn't include my "favourite," putting a heading for accessibility at a -10000 position, so screen readers would read it, but it didn't render on the screen.

Why would anyone downvote this? It's not a minority position.

Same here, but for 30 years. It's really no trouble.

I use Kiwi Browser. It is an open source fork of Chrome for Android that allows installing any extension. This is better than Chrome, which allows no extensions, and Firefox, which will only offer 200 or so. It also has its own ad blocking built in. Open source ftw.

I have used a keyed email address scheme since the mid 1990s (me_context@domain.org). 99% of the email I get now is sent to the same 5 addresses, in particular me_boingboing, me_linkedin. Those addresses were never on a web page. The keyed addresses I used for my company occasionally get spam, but it's generally on topic (offering relevant services, though still imo spam). I also have a gmail address that is a common first/last name but I don't send from that address, I get a lot of presumably guessed address spam there. This suggests quite a bit of organization in how spam works these days.

Thank you, I will take a look after current sprint.

There are already long running issues on this topic. https://github.com/oven-sh/bun/issues/2492

Thank you, it sounds pretty good, especially solving the modules problem. I will look into it when finished current sprint.