The computer revolution hasn’t happened yet (1997)
catonmat.net
catonmat.net
We are just starting to learn what computers are for. Hopefully, instead of being tools of stealing our focus and invigilation some day we will collectively figure out how to use computers so that we can live better lives.
Epidemiology ought to be able to easily tease out ideal "life best practises" - actual good advice given to you by your phone - backed by the mistakes of a million others in just your situation.
if keeping a daily diary helps people make better life decisions imagine what that would do?
There is already a lot of real value in ubiquitous computing, automated warehouses, supply chain optimizations, digital entertainment, digital government services in many countries, I could go on for couple of pages now.
Recommendations and tracking are just a side effect. FB and Google ads are not whole "digital economy" on their own.
Is technology about, or only about optimizing value in supply chains?
Maybe you thought about optimizing content delivery, but I was thinking about computer games, VR, music production, which are digital entertainment and have nothing to do with supply chain optimizations.
Imagine the power you’d have if you had people literally following orders based on a system of life that you’ve designed. And follow it they might if the perception of outcomes is positive.
Elo rankings (not ELO) was first adopted in 1960, which while technically overlapping with computing's timeframe, the two are unrelated. Certainly Elo was not invented to play online chess or for the sake of computer-aided chess by any stretch.
As of mid 2019 26 million air tickets and over 4 million high speed rail tickets were refused based on SCS. Also some personal information is openly published about people with low SCS.
There isn't one. We're simply the result of a bunch of lucky events and evolutionary forces.
I know some of my habits are awful (brushing teeth, exercise as of late, going on any walk outside) so I setup my phone to remind me every date at a specific time.
Can some of these be predetermined by an engineer? Yes! Can all of them? No.
Are not notifications intended by engineers to dictate a behavior? "To use my product" to what end various products are a means (Facebook, reddit, HN etc). I'd argue any product/system to which we join encourages us to follow it if we find it rewarding, regardless of positive outcomes (though surely the short term just look positive else what incentive exists to start?)
Every technological advance is usually both good and bad. Some are more good, some are more bad. Sometimes we gain a lot by it, and only lose a little, other times what we gain is not worth what we lost.
The problem with this idea is what makes life valuable the machine would not know. “Life’s best practices” is subjective, and therefore the machine could promote what one considers a best practice that another considers a worst practice.
The reverberations of the loudest (most magnified) permeate and shape ideology.
Hopefully the conductors get the frequency right for long term and long distance communication.
My personal takes:
For (1), I don't think it is self-evident. I see that the pleasure from the consumption of junk food and that from the mass media has some parallel, but tech has a far greater degree of freedom -- it's potential to truly transform human lives (for good and bad)
For (2), the food industry, although, by its nature, conservative, is going still going through evolution. Maybe fast food doesn't have to be junk food. Maybe nutritional balance can be achieved without compromising taste and cost.
I bought a action camera that is the size of a palm, with hardware 3D gimbal stabilization and can track faces and other objects. The video from which I can stream to my phone via bluetooth and it came to my doorstep 2 days after I ordered it online.
All of that involves computers of various size, sitting either in a miniature hardware in my hand or somewhere in a huge data center, and worked in tandem to provide an experience that is just mindblowing 20 years ago.
Was that camera somehow necessary for you to lead a good life or is it yet another distraction from it?
May I ask, what in a good life can not be classified as another distraction? Or are only human activities that someone from 1000 years ago can do pass the purity test?
The problem with the current trajectory of technology is that it seeks to strip us of autonomy. We are usually presented with means of consumption, rather than courses of action. I suspect that is what people mean when they discuss distraction.
I for one would much rather fill my life with pleasant distractions, instead of a Spartan utopian lifestyle that some other people tries to sell me as 'good life'.
Is the Internet contributing to the good in their lives? A report recently that found children who play social online games and connect with friends online while locked down have better mental health than those that don't.
https://www.theguardian.com/commentisfree/2020/nov/23/video-...
And the industry profits with that by applying psychological gimmicks in their games, the same ones applied by Vegas' casinos. The problem here is that they are applied on children that are unable to fully fathom the mid-long term consequences of their decisions and that keeps them playing and forgetting the external world. Their mental health is better until the day their parents take away their devices and send them to play outside. Suicide rates doubled over the years. Antidepressants use doubled over the years.
Social media by itself is not the problem. How big techs use them to profit (The Social Dilemma) is the problem.
I agree with you that this is an issue in any kids physical external activity. But the difference is that these activities are often supervised by an adult and in an event of a bullying the participants are removed and also often punished, and also the victim is helped and assisted. Note that I use the word "often" because I know that there are parents that encourage bullying behavior as ~self-defense~, which I also disagree heavily.
We don't have adults supervising every Fortnite lobby. Most children do not know how to use reporting tools to report improper behavior by another player. And while they suffer from bad words, psychological threats and such, they can't just turn off their devices, because the game always offer them a reward, a new skin, virtual coins, lootboxes, etc, so they keep playing, and therefore we have a loop.
Just yesterday I reported a player who cheated on FIFA, by exploiting a glitch that enabled him to play with a team full of superplayers and legends. The whole process took me almost 30 minutes. Sometimes I think that they don't want us to report bad behavior because the process itself is tiring.
Waiting 1 month to have something delivered is hardly what I call a good life. Let's celebrate the improvements where we can.
In lieu of that, I prefer my shipping delivered in 2 days vs. a month.
The leading COVID-19 vaccine involves a lot of computers. From the production of the mRNA strands to the setup of the supply chains to communications between scientists.
And there are few things that we need more than an effective COVID-19 vaccine in order to lead a good life right now.
The computerized chain that got GP his camera is similar the the one that is used in the making of the vaccine.
GP's camera is just icing on the cake.
I mean I'm guilty of that, I frequently browse reddit mindlessly. At least I'm saved the obvious ads - for now anyway, I'm sure they'll crack down on 3rd party clients soon.
The growth of the health food market - or at least some subcategories of it which are genuinely healthy - would be an example of this happening.
...assuming they're given a meaningful choice in the first place. Consumers choose from what's available on the market, not from some abstract space of possible products. Not only is technology too complicated for the mainstream audience (and products on the market do their best to make everything seem like "magic" that cannot be understood casually by non-specialists), the market doesn't really listen to user feedback anyway, preferring to divine everything from telemetry.
Greed, in general, motivates corporations to make money. Sometimes, providing people with what they demand is the easiest way to profit. Oftentimes, there are better ways. Like investing in marketing to make people demand what you want to provide them. Or scam them through convoluted business models.
Very true. However it’s easier for motivated corporations to steer people away from such information.
This can be done through both state-backed initiatives and non-governmental ones like Wikipedia, or speculatively, a public interest DAO that crowdfunds R&D. Gitcoin may be the first iterations of the latter.
Of course there are many cases where the consumer is underinformed, and harmed by what the market urges them to buy, and these need to be addressed.
This is just a specific narrow view of computers and how they are used by people in day to day life. The larger picture is that computers have impacted every single industry/domain and that impact have indirectly made lives better in some aspect.
In the end, computers are just tools, tools to automate tasks. What task humans want to automate depends on their needs/desires and so does the consequences of them.
In order to reach a superintelligence level we need to solve the automation of gathering, evaluation and improvement of knowledge first.
And by automation, I don't mean wikis or something you can find via google. Wikis and the web currently are built for humans, not machines. If we are able to automate the semantic knowledge behind it and transfer it to neural networks, we will have limitless capabilities.
As a kid, I spent countless hours in the underground trains staring out of the window into the moving darkness. Today, if I am on the train, I read interesting stories I like.
As an another example, apparently area about my birth town had all sorts of unique and interesting places to hike. I had no idea that those places even existed before internet - blogs and forums.
And the more unusual hobbies (think aluminum casting for example) - did you know how had was it to find good information before internet?
Please do not equate internet to facebook and twitter. It is so much more than that - you just have to invest a bit of effort instead of blaming it.
This seems a bit like saying that we're just starting to learn what engines are for. What we use computers for will continue to evolve and we will never arrive at an end state.
If people realized the value computing can give, consumer electronics will be useful for them. But they want instagram and snapchat instead.
A significant portion of the value created is not captured by economic statistics, as they mostly deal with measuring economic resources that are scarce enough to have market prices.
Abundant goods are free like air, and are widely distributed, and automation has made many classes of goods/services fall into this category, like many types of information (compare trying to find your way around a city 20 years ago, to today using Google Maps, or trying to read up on some obscure subject 30 years ago by visiting a library, versus finding the Wikipedia page on it via a 3 second Google search on your phone).
As you rightly said, I just can't think of a domain that's untouched by computers.
Even in the fundamental science it's now impossible to do anything without massive data crunching machines. LIGO detection of gravitational waves, the first black hole images, and the plummeting cost reduction of full DNA over last two decades are a great example of how computers have irreversibly changed the way science is done.
We want to believe that the revolution should be altruistic in nature. Bettering the human race through only "positive" methods - Star Trek style utopia if you will.
But reality shows that even with all this available computing power, it's still a finite resource. Any finite resource is capitalistic in nature and will be subject to the aggressive expansionist.
It needs its checks and balances like other resources.
If I leave out the bad parts then the rest is mostly what I need to use because of external demands or because of technology itself (ie. problems that only exist due to technology itself)
I don't need to be available on communicator. My team and my family expects me to be or I am marked as unhelpful. That wasn't a problem when there wasn't expectation to be constantly available.
I don't need to have Uber installed, but it seems to be the only way to get a cab where I live.
I don't need navigation, but roads and intersections got so complicated and I got so reliant that I would not be able to reach my destination without it.
Etc.
(Though I would say there's a gazillion ways we've gone forward. Videochat, youtube, twitter, and tiktok are all amazing things.)
Nothing in the history of life has changed fundamentally: Eat, reproduce, die. Your evaluation of the industry's products doesn't change the objective fact, that there is significant change enabled by computation.
Not sure we can build something fundamentally more capable than human brains. And if so, can we recognize the leap?
-“Wow Alan Kay is legendary.”
-“Can’t imagine life before OOP.”
-“I like how this YouTube video is annotated and I can scroll, see screenshots, and jump into the video where I want.”
-“I wish there were more video summaries like this.”
-“Generating summaries of videos like this is something tools like GPT-3 will do for us soon.”
HTML is a language to structure text with, that's perfectly fine to him. The problem is that the browser decides what that text should look like to you. Not only its aesthetics, but also how and where it gets displayed. If you want to visit webpages in a painting program ro find images, why not?
This makes a lot more sense if you watch one of his smalltalk demos, he draws a wheel and a car in a paint-like program, then tells the car to rotate as much as the wheel is rotated. He can rotate the wheel with his mouse.
Then he paints a pedal, and if clicked on it accelerates the car etc...
This is not a programming environment, but an operatjng system, you're free to make all your programs do everything
Coincidence or Simulation confirmed?
Or another way to look at it: during the N years I have lived, how much has the the world changed compared to the period of time between 1900 and 1900 + N? Though on reflection, I guess this is probably the sort of thought experiment that becomes more interesting as one gets older..
Additionally, many during the last 200 years probably thought they were in the most exciting period of technological innovation. When you factor this in the probability of existing during such a period seems much higher than coincidental.
I feel like it wouldn't be that hard for a person from today, especially someone who keeps up with scientific and technological progress, to quickly get up to speed with the world of 500 years from now... but I might be wrong.
A) you're just in an unlikely position
B) the population will rapidly shrink and never recover
C) this is the most "interesting" time in history, and there are so many simulations of it by the people of the far future that we are more likely to exist in this time.
And it's a daft argument because if you don't have a soul, you are the product of your environment. You couldn't be born as someone else, or somewhere else, or somewhen else, just like the River Amazon couldn't be on Mars or in Pangaea, because it's defined as "the thing in Brazil, currently". You couldn't be born in the Wild West because you are defined as "the child of your parents" and they weren't there, then. You didn't end up /in/ that meat body, you /are/ that meat body.
(And if you do have a soul, and they are randomly assigned to meat bodies, this argument is still like saying "roll two dice, the most likely combined outcome is a 7, I got two dots and one dot so that must be what 7 is")
Although echoing the opinions of many other commenters, I believe the early 20th century probably takes that crown.
The decades around 1900 were more innovative than what we've been experiencing the last decades.
And yes you can find people thinking the same thing way back in history, with the difference is that now it's been psychotically amplified by mass media, and it is actually annoying.
In popular press, and when people have things to sell, they do not include historical facts, because then they cannot drive the hype to new degrees, if they the include historical facts.
You may think that they have done their homework when they say "never before in history" but they have either never checked or deliberately ignored history.
I've had people I know, and the press swearing up and down that 'this time it's different' since the 90s when it comes to 'A.I' for instance. I'm sure older guys can go back even earlier and remember 'the impending A.I revolution'. Sweet money in that hyped narrative.
If you think that "this time it will be different" - congrats they got into your brain, and they do not care that they have done it just 10 years prior. With enough repetition of the same information, and the way the human memory works - it doesn't really matter.
You will help them every time by reliably assuming that someone else cared enough to check history before saying that something is "historical", "first time", "never before in millions of years of human civilization". They haven't, and it doesn't matter for their goals.
One of these times something will eventually happen. Until then PR money, clickbaits, VC money, startups.. the whole classical techno-utopian centrifuge.
Look yonder at yon contraption! A beast of iron and stream that can pull yet more buttloads [0] than a team of a hundred horses doth speed along tracks at a brisk pace of 80 miles every hour!
Yes, yes... quite right, Wigglesbarton. Truly we live in an age of wonders!
Well, I must away to beat my servants. Cheerio, old boy!
[0]: Actual unit of liquid measurement, eq. to 126 gal.
If you look at all of our youths, the biggest difference is between my dad and me, not between me and my son.
My dad grew up without TV, 1 person in the street had a radio. Later cars arrived.
When you look at my youth, we had computers, but internet arrived when I was 18.
My son watches TV, plays computer games, watches star wars, etc. Not that much has changed. Sure they have smartphones and social media, but the difference are details.
The biggest impact of computers is already passed us, and happened probably between 1970 and 2000
And it's just as likely to be a misleading perception today as it would have been at any earlier time.
How small we will look to them.
> In computer programming our basic building block has an associated time grain of less than a microsecond, but our program may take hours of computation time. I do not know of any other technology covering a ratio of 10^10 or more. ... This challenge, viz. the confrontation with the programming task, is so unique that this novel experience can teach us a lot about ourselves. It should deepen our understanding of the processes of design and creation, it should give us better control over the task of organizing our thoughts. If it did not do so, to my taste we should not deserve the computer at all! ("The Humble Programmer", EWD 340)
This. I think that the computer revolution would happen if programming teaches the rest of us more about our society and our mind, rather than computers and systems themselves.
Sounds interesting.
Often enough, the problem itself is easily solved. CS is all about how you approach the problem..
See also glamorous toolkit [2] for more immersive programming.
It's good to see that the way to really get it noticed is to present the contents in order as this guy has done.
I guess even Alan Kay didn't predict Google and Facebook...
In other words, I think we've long had the means, but capitalist underconsumption and the accumulation of technical debt holds us back.
Instead, we get silos, to better capture value. Users are consumers, and need only ingest what we feed them, to better keep them in check.
I'm not blaming anyone specifically here. Even OOP, and living cells tend to work the cell: keep your riches inside a membrane, and control every external interaction on your terms if you want to have it your way. Only... to some extent, as there is an equilibrium. Bacteria have plasmids. We at least have HTTP, for now (but it's not enough anymore due to captchas, js, etc).
The problem was summed up nicely in Stephen Diehl's "Near Future of Programming Languages" which I've posted as separately another comment (http://dev.stephendiehl.com/nearfuture.pdf). Here's an important slide:
"Where will the next great programming language come from?
Academia? NO. No incentive to do engineering. Those that do are committing career suicide. Funding is drying up for fundamental research.
Industry? NO. Can’t fund anything that doesn’t have a return beyond a fiscal quarter. Incrementalism doesn’t move things forward.
Hobbyists? NO. No economic means. Modern implementations require multiple FTE and decades.
Will we just be stuck in a local maxima of Java for next 50 years?"
I've personally attempted to create a sort of Manhattan Project of getting some top minds together in an off-site location to work on it in earnest. (I'm not joking. In 2018, I found a small 13-room hotel near Nice, France and made an attempt to get a bunch of us sequestered there for a year. Alan Kay liked the idea, by the way.)
None of that is happening. The sad truth is that the computer revolution isn’t even being worked on.
in later talks alan kay would say "the best way to predict the future is to prevent it" and i generally feel the same way...
it seems most commercial interests are interested in refining what already is, than building thind that are completely new (and who can blame them, its really risky to do so)
The irony is that everything is that by not investing in this, we've been eating our seed corn. There's been a dearth of innovation for decades and markets are desperate for growth but none is coming.
its intersting dilemma for sure... i think alan kay also said, the innovations for xerox park only cost a few tens of millions of dollars, with the economic impact of literally trillions...
then this book comes to mind as well: https://www.goodreads.com/book/show/6130130-my-years-with-xe...
> 00:43:15 Every object should be a virtual server and have a URL and an IP.
http://archive.cra.org/Activities/grand.challenges/kay.pdf
http://worrydream.com/refs/Kay%20-%20The%20Real%20Computer%2...
There's also a paper for those with access or a penchant for piracy
https://link.springer.com/chapter/10.100/978-3-322-89884-5_3
What I think we need is:
1) A global policy where computers are taught from scientific principles, with no tie-in to closed platforms or paradigms, throughout the education system
2) Ownership of data by data creators, with no platform tie-in, mandated standards for migration and access control
3) Engineers insisting on adherence to ethical standards for computing, such as doing no harm to humans. A case would be engineers actively participating in reviewing damage from tweaking algorithms for increased user engagement on crud platforms (time spent).
2010 https://news.ycombinator.com/item?id=1290498
2007 https://news.ycombinator.com/item?id=78349
2007 (a bit) https://news.ycombinator.com/item?id=75322
Not that much really...
We could have it interact with every citizen and act accordingly.
(I joke, but it couldn't be WORSE than the mess humans created.. right?)
Why should every object be a server. How does slapping a TCP and IP parser to every struct helps something. How does it relate to bacteria.
Yeah OK metaprogramming. The more a program use metaprogramming, the less readable it usually is.
I don't know. I am not Alan Kay and I didn't create OOP and Smalltalk so I guess I will stick to more practical things rather than philisophising about computer languages, and how objects will save everything. (I like Smalltalk for what it is.)
The point is that objects should be independent and communicate by messaging. The substrate doesn't really matter but both bacteria and computers on a network are examples of that scheme. The Actor model is the same idea developed for concurrency.
performance considerations aside, having every object (stateless or not) have an ip and run asyncronously (just like on the internet) allows you to scale out and replace the implementation at any time (just like on the internet)
if you are programming in the small (say c with structs) its hard to see the advantage of that, but if your programming system was "done right" you could have the same semantics and get both high performance and high scalability
one thing that alan kay says alot is, the internet is basically the only system to run for decades and still have all its parts changed and yet never taken down for maintanence (outside of living things that is, which is where the biological comparison comes in)
every object having an ip means, basically, the same thing, you could have a system run for decades in a distributed fasion and never need to be taken down for maintanence...
thats the idea as far as i understand it
I see. Going overboard with microservices has a long pedigree.
What a burn!