mojomark

FWIW, ChatGPT seems to do a perfectly fine job:

https://chatgpt.com/share/692a1119-e1d4-8009-95e3-259e52a19f...

I'm not sure I understand what the problem is.

Thanks. I do understand the value in holding debt, to my engineering brain it's analogous to a heat sink - without it, energy (money) has impetus to flow. I guess I don't have a good sense for what "the right" amount of debt is for a country, so to your point, I don't really have solid reason to assume we should even care about buying down our debt.

I think the root of my confusion was clarified by the other reply above - it's not that the US just has a stash of excess capital coming in and UBI is a way to spend it, but rather UBI is a way for a government to step in, and force it's ultra-rich citizens to put their money back into the economic engine at the base level... like an artificial mechanism to account for some percieved flaw in the natural capitalistic process.

Ah, thanks. So, the "universal" in UBI is a misnomer since everyone is society is not actually getting the stipend equally. Rather, it's a redistribution of funds from the rich to the poor. I would think a better name for it then would be a "Robinhood System" vice a "UBI".

Since this seems to be focussed on the US economy, can someone (...anyone), explain how a country that is many tens of trillions of dollars in debt can afford to pay a UBI in lieu of buying down national debt?

It feels like someone who can't pay their home mortgage deciding to make a donation to a favorite charity.

Well, that's not great. Gast turbine engines are notoriously inefficient compared to other electric power generation options (1). You use gas turbines when you need a lot of power, but limited volume (jets, warships). A server farm is not that.

1. https://www.nuclear-power.com/nuclear-engineering/thermodyna...

Ha, I wasn't going to read the article, but I had to after reading this comment. Yikes dude - I hope you don't ever happen upon my messy-ass/inefficient lab!

Didn't work for me. I asked for a man-made biomimetic submarine based on a blue whale swimming in the open ocean and instead got an actual blue whale flying above the ocean surface. Not very useful for me.

Didn't read. An LLM, of course, will think differently if it was trained natively on material written the "other" language (capturing the probablistic word associations inherent to the source material author's native language). However, if you train solely on English text and then simply translate the results to a different language, then no, there will be no difference. Seems pretty obvious.

I want to like this, but... While I do want generative AI for CAD, I really want it to generate initial files for me from scratch. That's where the bulk of the work is. Also, I'm not a building architect - my work is in mechanical and electrical systems. It wasn't readily apparent to me which domain this functions for.

The commercial vid was not very good IMO - I didn't get a sense of the actual product value-add at all. As an prospective customer, that's what matters to me, not terrible jokes.

I'm totally getting one of these for my office.

If I build industrial robotics behavior software where the constomers are other businesses and government organizations who may pay a one time up-front cost and then an annual sustainment license, can your AI personas actually accurately predict pricing in that domain (which is notoriously difficult) or is this just for consumer applications like meditation apps and the like?

Where are the e-mail-driven profits?

I use ChatGPT every day, to make myself more efficient (I get things done more quickly and with higher quality). Nobody really knows I use it.

The original comment is likely accurate regarding the benefit to ditectly trailing swimmers, but probably not trailing swimmers where shed vortices are stable in adjacent lanes where shed vortices interact chaotically.

Are you suggesting something like this could be used in an electroactive material that exhibits physical deformation to behave like an artificial muscle? (not suggesting that would be practical or efficient, but maybe?)

I'm inclined to concur with onlypassingthrough. If the resulting wake is similar to fish locomotion (e.g. thunniform or similar) vortices will shed off in a Karmen Vortex Street that spreads laterally with distance behind the swimmer (potentially into other lanes, and propulsive efficiency of propulsors are generally less efficient in turbulant vice laminar open-water flow... but not always, it can depend on the 'structure' [how chaotic] the flow is).

The magnitude of the energy in that turbulent wake will depend on how efficiently the oscillating fin interacts with water over time to produce forward thrust. The cool thing about oscillating foils as opposed to rotating thrusters, is that when the fin 'swoops' once it creates Vortex 'A' spinning clockwise, and when it 'swoops' back the result would be a Vortex 'B' spinning counterclockwise, and the two vortices will partially cancel out. That cancellation serves to recover energy from Vortex 'A' and the energy is transferred back into forward thrust.

In other words, fish tails create trails of contrarotating vortices and continually push off of them. It's like walking up a springy staircase, where each step you make, a little energy is recovered to bounce you up to the next step.

In theory, if you had a swimmer in front of you, generating a Karmen Vortex Street and not effectively canceling out those vortices, but instead just shedding vortices, you can use the energy from the swimmer in front of you to 'spring' yourself forward - barely using any energy yourself. Those complex hyrdodynamic relationships could be why some swimmers/flyers tend to fly in specific formations with other animals in their school/flock.

Bottom line, I would bet that any residual vortices that spread into adjacent swimming lanes will tend to interact chaotically and result in unstructured turbulance, which should yield less optimal swimming conditions for swimmers in those lanes.

This is cool, but I wonder what the implications are for the mutated molecule's characteristics that might result in microscopic material properties/behaviors.

Oh snap- this is awesome, well done! As I was going through this post I was thinking how awesome it would be to have a little robot trap set (snare, base drum, tom-toms, symbols, etc.) drummer that plays along to my pop music while sitting on my desk to give me a little show whenever I need to smile. Or maybe a little bongo player.

Super fun!

I don't know the answer to your question specifically, but in general, if someone tells you the resolution in terms of 'X by Y' pixels, that is unuseful information for an AR display. It drives me bananas that this isn't industry standard after all these years.

Instead, you need resolution in terms of Pixels Per Degree (PPD). And, to have any hope of viewing legible text on a virtual monitor projected via an AR HUD, like you can on a physical laptop monitor, you need at least 35PPD.

The actual article title uses the phrase "end-to-end encryption". The HN OP just missed the word.

I'm still honestly looking for an answer to my question below. The only thing this seems to do (to me) is add another minor step for a Hacker to read your email. Adding another authentication factor seems much more secure than this half-baked E2EE approach.

I don't get it.

Isn't the goal to prevent someone from hacking into your email account and reading your email? If someone already hacked I to Alice's email account, and pretends to be her to access the decryption key (which they can do because they can prove they have control of the account - which is all key access requires), then they can decrypt the message.

Seems pretty pointless, unless I'm missing a key point (perfectly possible)

I take it extremely personally, after all, I'm talking about voluntarily decreasing my own salary to achieve better work life balance in the present.

However, decreasing salary != decreased total compensation.

Increase equity shares + decrease salary = improved work-life balance now and much healthier retirement in the future.

Really weird that my comment was downvoted. No idea why.

I guess HN just isn't for me after 10 years. Peace out.

I put garbage in my compost bin. I get fertile soil out.

This is an interesting take, but I have a slightly different take. My ding dong company pays everyone including myself, salaries that are way too high. They gave ne a 30% raise I didn't even ask for or expect. We get a contract to do X, which say should take a team of 5 folks. However, we can only staff the contract with 3 folks because our salaries are too high. Now, we're all working 80hr weeks and stressed to the max because we really need more fresh minds on the challenge.

Now, we're in a contract lull, so some of us are only charging part time (75%), while actually working overtime.

It's bananas.

The solution seems obvious. Decrease salaries and increase company stock offerings. Then, if you and your colleagues care to work harder and deliver outstanding product, your stock value will rise over time and you'll be a.) more wealthy than taking 'dumb' money, and b.) more physically and mentally healthy.

Nothing lasts, that's why life is so great and interesting.

I just learned what a "zine" was, and also learned I've always had an affinity for them. One of my favorites is The Real Book [1].

1. https://en.wikipedia.org/wiki/Real_Book#:~:text=The%20Real%2...

LLM race may be over, but the AI race surely isn't. My baby seems to have grown into a fully functioning intelligence without reading the entire content of the internet. AI is not equivalent to LLMs, silly, silly child.

Well, I should clarify... today that's done on-prem. Of course if you had a secure network and host, LLM interaction could certainly be done in the cloud.

There are also other reasons to host LLM's locally. That is systems that leverage contextual reasoning power of LLMs, but are unable to readily access the internet. A planetary rover (e.g. Mars), for example.

Think larger - that is, governments. You simply cannot have people just typing whatever they want willy-nilly into cloud-based LLM's. For anything other than fully unclassified (and that is a lot - for any country) that must be done on-prem.

I would recommend all in this thread give "Cheaper, Faster, Better" by Tom Seyer a read.

This book has a liberal air, while I am a fiscal conservative republican (not a "Trumpian" 'flat-earther', nationalist, or handmade tale dreamer - never have been or will be... not all 'Republicans' are the same. These days I generally don't vote republican due to lack of sane options. F'ing painful and horrific).

[Just wanted to provide context for who was making this statement]

Disclosure aside, as anyone with two brain cells can see, the world has been snookered by the oil and gas industry and their kickback Trumpian-like Lackies for decades (old puppets same as the new). Seyer discusses the history of all of these topics, and additional similar ones, raised above fairly well (acid rain, LA smog, UK smog, etc.).

The biggest idea Seyer conveyed to me, that I wish was just a leaflet air-dropped accross the US is that Climate Change activists are preaching to the choir when they state that we need to reverse climate change because it's irreversibly harming the planet. They need to instead focus primarily on the human harm due to extreme weather (towns, societies, crop lands, god forbid ocean front property) that polution/carbon emissions are causing. I couldn't agree more.

It's like you have to point out how it's going to affect 'these people' directly, specifically, in their lifetime, for it to even register.

We are North American Scum. Unfortunately, common sense is a rare commodity these days. The internet gave too many selfish ding dongs a podium to preach to and 'unionize' other selfish ding dongs.

OK, this ding dong is off his podium. Sorry (but do check the bood out)

See [1]

1. https://forum.cktn.de/t/hacker-news-has-a-duplicate-submissi...