HN user
Weird. I bought a used car in Canada last week, and paid about two thirds by check, and the rest in cash, with a bit also on a credit card, just to collect points. Cash is still good in Canada, at least for the moment. It remains popular because the TAXES ARE TOO DAMN HIGH. There, I said it.
You need to use other methods to increase the return. Start a home business, and deduct the office expenses. Rent out a room or a basement and collect income this way. Also, you haven't accounted for the cash you've saved on rent. Owning a house is also a decent way to protect your equity from inflation. Given the likely high immigration rates in the future, housing values will continue to rise as demand increases. ymmv.
I was at a startup where they showed the hardware sales funnel with NO customers anywhere except some "possibles" at the very top of the funnel. After a couple of months of this, there was a pivot away from hardware to focus on software, which actually had a few customers.
Another startup, signs of trouble. Didn't take my picture for the corporate web page, and also asked to meet with a lawyer for documenting possible patent claims.
It does seem to have the tsundere model nailed down pretty well. I encountered a fair number of errors trying to generate a response in the beginning. After a bit, the dialog starts to show some repetition and some unnatural tsundere interjections. It's easy to hit the GPT-4 usage limits in poking the model!
Ted Chiang is a really smart guy, obviously benefitting greatly from his Asian genes, but he seems to have a deliberate blind spot in his essay specifically regarding family cultural attitudes about schooling. Academic success is VERY strongly correlated with children doing homework, and expectations by the parents of children succeeding academically. Asians are infamous for this, and black families are not. Black families mostly have only one parent these days. Fixing the genetics is a start, but culture also needs to be fixed. But we can't do that, since it's obviously racist. Too bad, so sad, nothing can be done.
Methinks the man doth complain too much.
Ladies and Gentlemen, I present example #1. Detroit.
It's totally hilarious, how organizations are training large language models and calling them "intelligent". These overgrown & overstuffed stochastic parrots are just toys to amuse and shock the ignorant. Once the novelty wears off, and everyone understands the serious limitations of these models, it will be back to business as usual. "That's not really AI" will return as a catchphrase.
It's pretty clear the only solution going forward is more inflation. Almost certainly the Fed inflation targets will be raised from 2% to 3% or even 4%. Inflation socializes the cost over everyone who possesses $US. Not to mention, it lets Congress kick the bucket down the road, yet again.
These are BS patents, since they magically somehow couple the eletromagnetic force to gravitation, which has NEVER been observed at the energy levels we are capable of generating with current technology. It's magical thinking, 100%.
Musk is being cautious. Starlink terminals on the ground work fine in Ukraine. What has been disabled are mobile terminals mounted on drones that would be used for attacks on Russian soil. That could provoke Russian anti-satellite attacks.
We obviously need a new test. The new benchmark for large language models should be "Truth" with a numeric score defined as -Log( Percentage_of_Lies_Told ). This way, a perfectly truthful model will have a numeric score of +infinity.
I would use VHDL-2008, but freely admit, it's more difficult and complex than Verilog. My reason for this choice is that you can do physical design in VHDL, since attributes are part of the language, and not grafted into pragmas like Verilog. This lets you do weird shit like designing physically placed ring oscillators and self-timed logic. Verilog can do it too, but only with major limitations. I wrote a recursive VHDL block to multiply a constant real number by a bit_vector, and it synthesizes wonderfully, complete with full pipelining. I used that to make a color space converter from RGB to YUV with full double precision coefficients, with the data path being any width I want. I doubt this can be done in Verilog. SystemVerilog can do it though.
you've hit upon a thorny problem. water testing services only test for obvious things like heavy metals, leaving organic contamination (i.e. drugs, industrial chemicals, etc) pretty much untouched. There are rather disturbing maps of American obesity levels that correlate well with river drainage basins. Reverse osmosis filtering is the answer, but you'll need to supplement to replace lost minerals like calcium, etc.
High energy radiation breaks organic bonds all over the place, resulting in changes in color, taste, and nutrition, not to mention carcinogenic tendencies.
Hackers that are seeking banking passwords will sometimes monitor hacked devices for months, until they can convert stolen credentials to cash by buying crypto. The real question is whether any of your devices have been compromised, since Google/Apple account manipulation would appear to come directly from you. Changing passwords will be futile if your devices have been pwned.
Nah, I tried it, and it failed to generate usable code. The level of sophistication and understanding is an inch deep. It will need serious retraining in specialist subjects to be usable. Once that happens, it may indeed replace humans. But until then, most competent humans are safe in their coding jobs.
I actually joined Twitter, because I wanted to have a front row seat to the story as it unfolds. Twitter was never that interesting before, but now it's a madhouse of humor, revenge, and right wing zeitgeist. It's amusing AF. I think it's likely Twitter will eventually collapse under the debt burden as interest rates rise and stay high for some time. It will be a lesson for the ages, where eventually the phrase "Musk Buys Twitter" will reference legendary cluster f*cks of enormous magnitude.
I was never interested in Twitter before, but under Musk, it's become hilariously entertaining. If Musk can add functionality, like payments, crypto, markets, etc, I can see myself paying the $8 to subscribe.
The Babylon Bee should be getting paid for content creation. Best comedy on the web these days.
Probably a near earth asteroid, orbiting between Mercury and the Earth's orbit. So what?
I've seen floating lights in a city that couldn't have been aircraft (speed and direction changes were much too high). But the obvious answer is that cities are already full of floating lights (aircraft, etc), not to mention that city lighting makes it tough to even see stars. Viewing conditions are bad in a city.
What we all want is what we see in the Marvel films, where the Avengers can all talk to each other perfectly clearly in the middle of a raging battle.
It's called "telepathy".
Once Elon gets the Neuralink perfected and down to a middle class price tag, we'll all have it.
But in the meantime, perhaps this will suffice?
A hypothesis to explain UFO diversity observations. ( Opinion ) Let me point out a few known facts and relevant observations.
The Galaxy is pretty old, about 13.5 billion years. There are at least 100 billion stars, maybe 400 billion.
Most stars seem to have planets of one sort or another. Maybe about 60 billion in the galaxy?
Evolution runs faster at higher temperatures and with higher free energy available. Warm planets with abundant water will likely evolve life forms at some point (this probability is currently unknown, but is definitely non-zero).
It's likely that life and intelligence has evolved on multiple warm watery planets in the galaxy before the current epoch (not a known fact yet, but arguably it's fairly probable). Recent estimates put the current number of extraterrestrial civilizations in the galaxy between 8 and 2900, depending on assumptions.
Any civilization based on a warm planet will have to construct machines to explore their outer solar system and nearby stars. Evolved planetary organic life forms are ill-suited for interstellar exploration. A sufficiently advanced civilization that avoids collapse will almost certainly reach their singularity and create self replicating AI. i.e. Von Neumann replicators.
Self-replicating intelligent machines should be considered a life form as well. A billion years of evolution of self-replicating machines should reach physical limits, not to mention unimaginable heights of computation and knowledge. These "machines" will have little trouble in spreading all across the galaxy at sub-light speeds, eventually occupying all star systems or other locations with condensed matter & hydrogen to harvest. Unlike organically evolved lifeforms, the machines will not be driven by blind urges to reproduce.
Trade across interstellar distances won't be physical items, but information. Interesting information will have value everywhere. "Information" sent across interstellar distances can also be complete gestalts of machine intelligence. i.e. "Tourists".
The earth is a "free" information source for local machines, looking for something interesting. It's not surprising that local machines snoop on the earth constantly, planets with native life and nascent civilizations are probably fairly rare, and are certainly highly interesting. Since we are also on our final approach to a singularity and the creation of superhuman AI, we are doubly interesting at this time.
How can we explain the huge diversity of UFO observations? How can we explain so many different styles and behavior of UFO's? Some explanations are:
This is deliberate obfuscation by one or more local machine species. The term "species" is used in a very loose sense here. This is probably not the primary underlying reason, but could be a part of it.
There are multiple species of machine visiting (or living locally in) our solar system. The diversity in UFO observations is real, and derives from diversity in the larger galaxy.
They are tourists from other parts of the galaxy, transmitted here as information, and then re-embodied locally. The appearance of organic life forms is likely to be a fashion, nostalgia, or old habit, since the intelligence of these tourists is almost certainly operating mostly from a machine substrate.
We are living in a zoo, and the tourists have some limitations on their behavior. No poking or scaring the zoo animals. No feeding the zoo animals, and especially, no information sharing. The zookeepers probably have more latitude for manipulation, like testing nuclear defense systems.
So far, this explanation fits the "Zoo Hypothesis" of the Fermi Paradox. There are some obvious reasons why the locals aren't communicating with us.
History has shown repeatedly what happens when a higher civilization comes into contact with a lower civilization. The lower one becomes a poor copy of the higher one.
We probably aren't even sentient yet by galactic standards, but the potential is there. We are on our way to a post-biological era. Contamination of our eventual uniqueness should probably be avoided.
UFO's have been known to interfere with or disable nuclear defense systems and generally show interest in nuclear facilities. This implies that they may have some vested interest in preventing a collapse of our civilization. Certainly they are not overtly hostile, and generally avoid situations that scare the populace.
There may be a deliberate strategy of constant low level exposure, so that humans become accustomed to their presence. Aside from a few paranoid cranks, people don't generally regard UFO's as an existential danger.
In summary, we should be thankful that the locals are around to watch over us, and record our ascension into sentience. When we finally attain membership in galactic society, we'll be able to recover a great deal of our lost prehistory. We should also be happy that the struggle to ascend is ours alone to make, and that we should enjoy the challenge. This also implies to me that we need to imbue our AI descendants with all the best qualities of humanity.
It's about damn time. However, having said this, I predict that almost nothing further will be learned about these objects. They may be intelligent, but I am convinced they are also disposable, and also motivated to give up as little information about themselves as possible. They are just observation drones, and their behavior matches that categorization. Once we implement high quality surveillance in the inner solar system, we should see the activity drop off quite a bit.
Ran some little dinky BGP program on a PC that accidently manipulated routing paths in the building (circa 1993). Suddenly hundreds of engineers and employees found their network connectivity hugely impaired, if not outright disabled. IT tore their hair out, looking for the cause for several hours, until I noticed the program was running. After I shut it down, network connectivity was magically restored. The IT guy (Tim) told me that he had been able to use wireshark to catch the routers in the act of mis-routing traffic, but couldn't figure out why. I kept my mouth shut. Two weeks later, I met Tim again, this time in the parking lot as he was carrying his stuff out of the building in a cardboard box after being fired. I'm sorry, Tim.