One time, I was shopping for a specific item. I couldn’t find it on ebay, Amazon, walmart or etsy. Then I went to some smaller retail site (they also didn’t have it), and an Amazon ad for that item popped up. I clicked the ad, and it took me to the item page.

Amazon search (at least at the time) was so ineffective that I couldn’t find it, while their ad data gathering was so complete they knew that I wanted that specific thing.

Inflation is wild. Just a few decades ago, you could get this kind of thing for just an arm and a leg.

They were going to fire up the LHC to open a dimension, which can only happen when the moon is casting a shadow 6000 km away. Then aliens… or something.

The problem when photon containment breaks like this is that we can never be 100% sure which photons were SUPPOSED to be there, and which ones leaked out. We’ll need a dedicated team of particle physicists with very small tweazers to have any hope of sorting out this mess.

So you’re saying my proposed imperial units depend on where you are, and who is using them, for what purpose? That just sells me on them as imperial units even more. :)

Thank you for the details.

From smallest to biggest:

Bits (basic unit)

Bytes (8:1 reduction)

Words (4:1 reduction)

KiB (32:1 reduction)

MiB (1024:1)

GiB (1024:1)

TiB (1024:1)

PiB (1024:1)

A normal amount of porn (237:1)

I counted twice. It didn’t match.

Cheddar saltine crackers. Premium made them, but I think it was only for a year or two.

What we have done is invented massive, automatic, no holds barred pattern recognition machines. LLMs use detected patterns in text to respond to questions. Image recognition is pattern recognition, with some of those patterns named things (like “cat”, or “book”). Image generation is a little different, but basically just flips the image recognition on its head, and edits images to look more like the patterns that it was taught to recognize.

This can all do some cool stuff. There are some very helpful outcomes. It’s also (automatically, ruthlessly, and unknowingly) internalizing biases, preferences, attitudes and behaviors from the billion plus humans on the internet, and perpetuating them in all sorts of ways, some of which we don’t even know to look for.

This makes its potential applications in medicine rather terrifying. Do thousands of doctors all think women are lying about their symptoms? Well, now your AI does too. Do thousands of doctors suggest more expensive treatments for some groups, and less expensive for others? AI can find that pattern.

This is also true in law (I know there’s supposed to be no systemic bias in our court systems, but AI can find those patterns, too), engineering (any guesses how human engineers change their safety practices based on the area a bridge or dam will be installed in? AI will find out for us), etc, etc.

The thing that makes AI bad for some use cases is that it never knows which patterns it is supposed to find, and which ones it isn’t supposed to find. Until we have better tools to tell it not to notice some of these things, and to scrub away a lot of the randomness that’s left behind inside popular models, there’s severe constraints on what it should be doing.

That estimate is based on assuming that the ratio of matter to light output is the same between galaxies 10 billion years apart in age. The high light output of these young galaxies could also be supermassive stars that burn out very quickly, larger stars typically forming faster than smaller stars, or many other things.

Blindly assuming a linear relationship between two things, then extrapolating is how you get the Windows loading bar circa 2000.

Separately, but just as big a potential issue, the data itself may be incorrect. Previous galaxies measured at extreme redshift values were remeasured, and found to have less extreme values. This can be as simple as there aren’t that many photons from these galaxies reaching us, so a short measurement period might not be enough to get an accurate picture.

As the President and CEO of a fortune 500, and a neurosurgeon that does rocket surgery as a side project, there are many people relying on me to be an upstanding member of the community at all times.

In reality, most of the ways I misrepresent myself are to obscure my identity, and mostly it’s by leaving things out.

When we get to this Saturday, next Saturday moves a week beyond our grasp. In some sense, when we die, our own timeline ends, and we can finally arrive at next Saturday for the first time.

Because the longest lives are around 110 years, and children start speaking around 3, the furthest in the future “next Saturday” can be is abouy 107 years.

Yeah, but then you have to limit your fashion selection! Imagine restricting yourself to wearing a hat over some silly fear of death.

I appreciate numbers.

I almost get 100,000 pis. Almost, but not quite :'(

I’m… uhh. I think I’m screwed here.

There’s a crucial distinction between someone that wants to have sex, but cannot, and someone that chooses to identify as that. To really become an “incel” in the negative sense, you lose the desire to have sex because being denied sexual contact by others is part of your identity now.

People that merely don’t find others that are sexually interested in them can do things to help themselves, learn better grooming habits, dress nicer, practice approaching and talking to people, etc. Someone that has adopted the identity of “incel” can only help themselves by changing their perception away from the toxic void they found.

My great uncle said he hoped for a quiet death.

If you can prompt it, “Write a book about Harry Potter” and get a book about a boy wizard back, that’s almost certainly legally wrong. If you prompt it with 90% of an article, and it writes a pretty similar final 10%… not so much. Until full conversations are available, I don’t really trust either of these parties, especially in the context of a lawsuit.

I don’t think anyone will actually make it, but it would be cool to have an arrangement of accelerometers and microphones that you can put on the side of a packaged gift, shake it, and get a guess about what it is.

A harvesting robot that can tell how many days from ripe an avocado is, so the grocery store can have like… “ripe today” avocados, “ripe tomorrow” avocados, “ripe in 2 days” avocados. They’d come in small cardboard boxes, and they could just shift the boxes or signs over by one each day, and have more boxes if they get avocado deliveries less often.

Machine learning clothing/hairstyle/general fashion advice would be neat, but probably too open to manipulation to sell certain brands to be practical.

Tools to help developers put houses at the best spot on a lot, for things like water mitigation, tree safety, garden space in good sunlight, wind noise, and privacy.

Search tools that aren’t terrible on shopping sites, and news sites, and research journals and things. The days of “we asked Google to do it for us” being good enough are long over.