Jeri Ellsworth and the Robot Uprising of 2038 | Hacker News Reader

Jeri Ellsworth and the Robot Uprising of 2038

blog.robertelder.org

199 pointsrobertelder65 comments

"Well, magnetic core memory is the only data storage format that is robust enough to withstand high-radiation environments. Jeri is clearly interested in magnetic logic and memory because it is the only computing platform that will be able to survive the first wave of nuclear blasts that will unavoidably come from the beginning of the third great world war. "

Erm, this premise is factually untrue though. A lot of next generation resistive RAM devices, especially OxRAMs, have been demonstrated to be rather rad hard, making them good candidates for future space electronics platforms or.. all the other attendant apocalyptic scenarios.

Radiation hardness is different than sensitivity to EMP. It is the eddy currents from an EMPT that build up in and burn out small traces in micro electronics.

But magnetic cores are resistant to the M in EMP?

I'll offer a wild guess - perhaps they are only temporarily affected (i.e. mem-wiped) and function as normal after device reset, being made of iron, in contrast to semiconductor doping materials being hard-killed by the emp.

Drum memory [1] and hard disks [2] have been used in nuclear tipped missiles, which are supposed to operate in an environment where EMPs are expected. Both types use magnetism for storing data.

[1] https://en.wikipedia.org/wiki/ASC-15 (Titan II)

[2] https://en.wikipedia.org/wiki/D-17B (Minuteman I) (note that the hard drive was used as RAM)

I have not read the article, but I find the choice of date interesting. 2038 is unix rollover time [1]. It does not appear to be mentioned in the article so it might be a coincidence.

[1]: https://en.m.wikipedia.org/wiki/Year_2038_problem

---Edit---

> but the meek will inherit the earth when judgement day begins on the 19th of January 2038 at 03:14:07 UTC.

I stand corrected, Unix rollover time is exactly why this date was chosen. It must have broken AI safety protocols in the robots which turned them into killing machines. Cool, I will read this article tonight.

Wow that was... Extremely weird. Is this supposed to be some kind of tribute?

I think the idea is to use the idea of future robot dictatorship as a spurn to learning DIY transistor manufacture.

Didn't read the whole thing but definitely cyberpunk

I've met Jeri. I'm pretty sure this guy is onto something.

I only wish she'd get the clone machine working ASAP.

Weird article. Author isn't technical enough to get how Jeri thinks. I've met her. She sees the physical world in terms of first principles, like Feynman. She built a short range radar from a satellite downlink dish by turning one of the transistors around. I would never have thought you could use a transistor from a low-noise amp to produce enough power to emit a usable signal. I'd be thinking "small-signal transistor" vs "power transistor". But it's an amp, it's got gain, it's fast enough, and she made it work.

If you like learning about low power or messy signal stuff, there's some research at UW several years back about using e.g. a combination of em interference from human bodies and normal house electrical wiring/lightbulbs as motion detectors.

That sort of stuff is really cool, and I'm very impressed by people like Jeri who can actually make it work.

>Author isn't technical enough to get how Jeri thinks.

You know, the author could be telling... a joke.

FYI: https://en.m.wikipedia.org/wiki/Jeri_Ellsworth

There is a youtube talk linked in the articles references which is ... just watch it. I am always amazed by these stories - especially as I tried the opening a store in mid nineties flogging computers and internet cafes.

I for one really liked that Blade Runner reference with the repetition of "interlinked". My initial vision of the post's future was of The Terminator, but that reference made me consider a more interesting future with different kinds of artificial beings.

What makes core memory resistant to EMP?

Wouldn't the pulse change the polarizations stored in the iron cores?

> The highly irradiated port-war landscape will be unable to make use of standard silicon computer memory for centuries to come.

So we really couldn't use any silicon that was hit to make new semiconductors?

The author’s argument is that a highly irradiated landscape would be too energetic for silicon based memory to work as it’s now designed. I’m not an EE but I’ve spoken with some engineers of early silicon systems which had to account for high energy cosmic rays flipping the bits within RAM in a random fashion. My guess is that using a solution like Hamming codes would not work when retrieving arbitrary data at a large scale.

Ecc does work for radiation hardening of storage. You can still get logic upsets in transit, between storage elements.

You can either design your cells around it (high power ) or you can use a variety of redundant logic and monitors.

Off topic, but with smaller lithography, ANY subatomic particle of sufficient energy becomes a problem: https://www.xilinx.com/support/documentation/white_papers/wp...

I've spent months chasing random crashes on another vendor's device which turned out to be solder ball alpha emission issues.

"My dad had bought me a 101 electronics kit from radio shack..."

Proud to say I had the same kit. Santa is the man.

>"I ended up dropping out of high school when I was racing cars because I was making so much money."[3] "I started seeing how far I could push the rules. One of the things I built was a traction control system for my race car. I built a single-board 6502 computer (wire-wrap) and I measured the front-wheel spin by putting a hall sensor on it, and I measured the engine RPM." "I had a rev limiter in the car and it would just tell the rev limiter you're over-revving if the back tires were spinning more than the front tires." "Then, I started dominating, just completely dominating.

Sounds a lot like cheating......disruption

Finding loopholes or under-specified gray areas in rules to gain an advantage is the secondary activity, besides actually racing, in almost all motorsports.

Crazy genius. This is an example of someone born to do something.

I know that's just supposed to be a complement, but sometimes that kind of a comment contributes to people thinking there's such a thing as a "math person" (or a "non math person"). Having comments like that thrown around in the aether makes it seem like technical subjects have "you either have it or you don't" characteristics, which I think most people on HN know is BS. I prefer to emphasize hard work over any kind of genetic advantage.

Yet there is a difference in people whether they start creative something or not. Whether you prefer something or not, doesn't make genetic differences in mental stuff less real.

That is how people are. We are not identical, interchangeable cogs.

Most people can be trained to do simple things in most any field. But brilliance is innate. Its not learnt and can't be made through hard work. Some people have analytical minds some don't.

I'm pretty good at what I do, and when people tell me they're jealous of my talent, and kind of hint that it's innate, I get a bit insulted. I had to work a shitton to get where I am, and comments like that makes it sound like I had it easy.

The reality is probably a mix of both.

There are people who could work harder than you worked and not get to the point you are at.

> I had to work a shitton to get where I am, and comments like that makes it sound like I had it easy.

Not at all, but you had it possible. There are plenty of people who are just not equipped to do whatever you do and will never be halfway competent, let alone pretty good at it, no matter how much training they receive.

We're all good at different stuff in different amounts. There's probably some stuff they can do which you can't, too. That's OK.

It's not that you don't have to work hard if you have talent. It's that if you don't have the talent, no amount of hard work is going to make up for it.

There are so many people in the world, so much competition. That to be "top" performer in whatever field you have to 1) have innate talent, 2) work your ass off 3) get lucky.

I would say opportunity + curiosity + propensity is the formula and we know what curiosity and propensity are not chosen things. The other caveat is just anomalous ability and although rare we have strong evidence of these people existing. There's nothing wrong with that. I don't have the baggage that you speak but understandable. What you mean is applicable for the average folks, me included.

I get what you are trying say. It is just that most of society doesn't care if you are smart, creative or trying to do anything special when you are trying to get an education, apartment or anything else on your own terms. It is only once you succeed that people are quick to call you all those things before going back to enforcing the same rules you barely survived. It is sort of like an honorary degree.

There is something to be said for that, 20/20 hindsight is easy to classify and mistakingly so bringing reasons for things unknown or impossible to know ahead of time is silly. I get that too.

> It is only once you succeed that people are quick to call you [smart, creative, etc] before going back to enforcing the same rules you barely survived.

This very eloquently sums up most of my experience with society. If you succeed, it's 100% because of talent or privilige or dumb luck or not having to work a part-time job to feed yourself/your younger siblings/your ravenous creditors, and 0% because of any actual effort on your part. If you fail, it's clearly because you're a lazy good-for-nothing welfare-sucking leech, and any attempt to point out other reasons is just excuse-making.

"If we succeed, no one will remember. If we fail, no one will forget."

(Incidentally, while trying to find the source for the above, I discovered that Google considers it a "inspirational quot[e] about overcoming failure".)

Most people can't be Jeri any more than they could be Michael Jordan. 10,000 hours of practice wouldn't be enough. A million might, if you somehow managed to combine it with a foolproof nootropics regimen or something equally fictional.

You might not like it, but that's the way it is. All (wo)men are not created equal. Whoever told you they were was lying, or at best misinformed.

My experience is that "you either have it or you don't" is more accurate than not.

> I'm pretty good at what I do, and when people tell me they're jealous of my talent, and kind of hint that it's innate, I get a bit insulted. I had to work a shitton to get where I am, and comments like that makes it sound like I had it easy.

This is a common sentiment, but I don't find it convincing. The discussion here is about nature vs. nurture. There's a lot of data points on that subject, and, from what I've seen, they mostly support nature (twin studies, intelligence being 60-80% heritable, etc).

The fact that you like to think it's mostly nurture isn't really relevant. I'd like to live in a world where ability and virtue were absolutely correlated and everyone got what he or she deserves based on how much good that person does in the world. But I don't think we live in that world.

But for one guy, here is a completely unimpressive family[0]

But which studies show those 60%-80% numbers? A 20% spread is a lot but not surprising since the results change completely with the definition of intelligence which is easy to redefine since it is just a social construct.

[0]https://en.wikipedia.org/wiki/Einstein_family

>My experience is that "you either have it or you don't" is more accurate than not.

Maybe that's correct, but is it more useful than the nurture view?

It's a debate; I mean, clearly, you don't want to spend a lot of time trying to get better at things you will never get better at, but sometimes it's not easy to tell you will be good at a thing until you spend a lot of time on prerequisites.

This is the argument that even if it's true that nature is more important than nurture, it is sometimes more useful to believe that nurture is more important, and that you can learn things.

I think there are a fair number of studies showing that people with a 'growth mindset' as they say, who believe it's about hard work and not just innate ability, tend to do better than people who believe it's fixed.[1]

I mean, in the usual case, of course, people who are smarter are more likely to think than anyone can learn things, 'cause their experience is that learning things is easier, and the way work and education is segmented, quite often people are put near others of their ability. but... I think a lot of these studies control for that.

[1]https://news.stanford.edu/news/2007/february7/dweck-020707.h...

My own personal view is that some attributes are mostly fixed, and some are not, and it's pretty hard to tell which is which.

Sometimes you need to struggle for a long time through stuff you aren't any good at to get to something you are good at. I love reading and read very quickly, with above average, but less impressive compared to speed, comprehension.

But it took me longer than usual to learn to read. I assume because I'm terrible at memorization; and to learn to read, you need to memorize enough words that you can start puzzling out the words you don't know from context. I would not have done well if I had given up on reading because I was inherently bad at one of the prerequisites.

On the other hand, I spent a lot of time trying to run a business... and that turned out to be something I didn't get good at after a decade of trying; so yeah, sometimes it's best to give up early, but it can be really hard to predict which side of that equation a particular problem is on.

A better comparison to reading is writing. My handwriting is terrible. Like not doctor-terrible; there's at least dignity in that. It's a grade-schooler's block letters. It's terrible. And I spent so much wasted time and effort on it.

I mean, I've had access to a computer since the mid-80s, so I could write, I just couldn't hand write. And it turns out? nobody cares about my handwriting anymore. It's not that useful when you have portable computers. (In the mid '90s I took a tandy TRS-80 model 100 to high school. Such a nice keyboard)

I argued and fought with my parents who made me practice handwriting, arguing even in the early '90s that it was an obsolete skill; but they sat with me for hours a day, making me copy letters. Just like they did earlier, when they were making me learn how to read.

But reading, well, it went from a struggle to make me practice to the thing I got in trouble for doing when I was supposed to be doing other things almost overnight. It was like I finally memorized enough of the words to figure out young adult fiction, and a whole new world opened up to me.

Handwriting never got to that point, even though there was a lot more struggle involved. To this day, I can't write coherently with a pencil and paper for more than a sentence or two; it takes too much focus to make the letters, and the thoughts about the sentence or paragraph evaporate.

How much of this is my own motivation? I do remember arguing that handwriting was useless 'cause I could type, and that was better. If I had access to a state of the art screen reader (they existed at the time, and I think were okay?) I maybe would have made the same arguments about reading. Would that have made it harder for me to actually learn to read?

As a most extreme counterargument I raise you the NBA players. They got it, we don't. No amount of training can make up for it if you don't have it.

Yes this is a clear cut argument. The fact that the brain is hidden under the skull allows people to make statements that people's brains are all the same and its just effort after that. I think generally they are correct but there are outliers, most likely similar in proportion in terms of the traits that make NBA players in rarity in terms of intellectual gifts. I have met only 2 genius people in my life and the difference was so strikingly obvious that there was no doubt. One was a child prodigy in music and the other was a brilliant programmer that did string theory in high school. I also played basketball in my youth with a kid that made us all look like goofy idiots playing against him (but later drugs did him in) and swam with a swimmer that could lap everyone in the whole league but I do not know what came of him. It is a gift to see these people and undeniable when you experience this but for the vast majority of us, hard work and sheer will makes a huge difference in our own outcome.

You've never tried to tutor a computer class, have you? There are absolutely people who naturally 'get it' and people who never will, along with a large middle ground of people who don't 'get it' innately but are capable of learning.

This tabula rasa dogma might help determine how to treat people from a human perspective, but when trying to help people choose a career path it's terrible.

She did, she founded Technical Illusions with a fellow Valve employee when they were fired from Valve. It was a AR bid. They raised millions of dollars but eventually shut down.

One awesome project was called CastAR https://youtu.be/AOI5UW9khoQ. It didn't take off like porn or anything, but that doesn't mean it wasn't a really cool technology.

After the company folded, her and a few others bought all the ip and prototypes for cheap and are trying to restart it correctly.

I found a new person to follow! From this blogpost and her wikipedia page I'm impressed.

I wonder though -- is the fear of the Robot Uprising of 2038 founded, to her, or is that applied drama by the blog author?

I don't know her personally, but I've followed her for a while and I've never heard her mention anything of the sort. Also, the author didn't mention anything about her work with augmented reality, I guess it didn't fit the narrative. To be fair though, I only read the first 1/4 and skimmed the rest b/c I couldn't stand the robot uprising slant. Can't she just like cool old tech without being a cyberpunk pepper?

If you're interested, she has done some excellent interviews on the amp hour podcast. The episode referencing her semiconductor making experiments: https://theamphour.com/the-amp-hour-52-carnassial-chip-chemi... All episodes: https://theamphour.com/?s=jeri

(edit: copy/paste error)

Highly recommend Jeri's youtube channel

I have a bit more positive look on the future: Jan-2038 is just before my retirement. I can make some money with old computer knowledge while the post-post millennial generation does not even know what the inside of an old operating system looks like.

Your old computer knowledge will be about the same worth as what income that can be generated today having extensive knowledge of Multi-User Dungeon forum software.

The world does not run on MUDs. The world does run on 30 year old OSes.

Perhaps, perhaps not. I’ve seen things go both ways.

On the one hand, circa 2005 I made an audio recording at 16-bit/44k samples per second, of my graphics lecturer saying “40 joints, 24 samples per second, 16 bits per sample is a huge amount of data. A megabyte after only five minutes!” (My hard drive being 40 gig back then).

On the other hand, when I graduated I ended up seeing several of the algorithms he taught us in use in workplace that did a lot of highly optimised graphics processing.

So... incredibly lucrative if Ready Player One is to be believed.

MUDs were advanced data processing systems. As a kid I wrapped my head around one with a 100k line code base. That has led to quite a successful career and I still use a lot of the same techniques I learned.

That's cool, yeah. Must've seen some interesting coding practices in that code base

https://github.com/smaugmuds/_smaug_/tree/master/src

I still do C so it isn't so far removed.

brilliant,

but for

>Triboluminescence with household chemicals - Keep your eye on what happens when the battery falls over in this video.

I watched it over a couple of times, but never actually catch the battery fall over, the subject matter is constantlly grabbing my attention. at what time does it fall over?

But now that she's on the internet the robots will know to find her after the uprising.

Jeri likely would find a way to defeat the robots by building a machine gun that shoots pinballs and is controlled by a brainwave interface to a SID chip.

Catch 22. The killer robots will be quite proficient in tracking electromagnetic radiation.

The robot uprising will begin and I predict the battle turns at 12am Jan 1st, UTC, as the 32 bit system clocks overflow in many of the small micros controlling subsystems.

starts with a typo

> Jeri Ellsworth hates this article

probably used in jest - but playing along - probably because premise of article is heavy on summoning up ethos & pathos to generate interest whereas, from what I can tell of Jeri's style, she is more centered on direct logos.

in a way, the way OP uses ethos/pathos method is somewhat deceptive/dishonest, because it's trying to get you interested in source material for reasons that are entirely make-believe.

Reply on news.ycombinator.com