The introduction of writing and agriculture were watershed events for humanity, and are forever interlinked. People have been telling stories – made up and otherwise – and sharing information about the real world for perhaps hundreds of thousands of years as they roamed the Earth looking for food and shelter and avoiding enemies and threats. But when they settled down, we think to make alcohol – which is fun, which is a great pain killer, and which is still the reason a lot of us got created in the first place.
Bread came after alcohol, people. Anyone who knows people knows that.
Suddenly, there was a surplus of food that created wealth, and there was a need for a hierarchy to track that wealth and to maintain the human-powered machine that created it on a scale that our nomadic predecessors could not have imagined.
And here we are, some tens of thousands of years later, technology advance after advance, tiny steps at a time, at a point that any civilization eventually reaches if they don’t blow themselves to Kingdom Come with nuclear weapons: The rise of the machines. It is, perhaps, speaking abstractly about a Universe so vast, happening millions of times to zillions of worlds in various states of evolution, in the immediate vicinity of the Milk Way galaxy. Or, perhaps we are the only one. Perhaps the Universe is designed with vast distances in space and time specifically to make it hard for us to know, perhaps specifically to keep us from interfering with each other if there are indeed many worlds. Perhaps a Universe is a big ensemble model, changing initial conditions here and there to see how it all plays out. Who can know for sure?
In this sense, we are on our own, and each world either gets through it or doesn’t. What we can know for sure is that this is happening to us now.
By their very nature, writers are hoarders of information and experience. And we were always writers, from the womb, and long before we ever became journalists in the tech field. Nicole and I are the same kind of animal, the only other ones of our kind we have ever found, and we found each other because of this fascinating business of high performance computing. And because we love story, a decanting and recanting of what others have experienced for the sake of you, we are concerned about what is happening now in a way that we have never felt before.
Collectively, we have seen quite a bit of change not only in systems, but in the value of information, of experience, and of creation.
Back in the days before the commercial Internet, in the wake of the Wall Street crash of 1987 that tipped us into recession a year and a half later, I started my career in boutique IT newsletter publishing. We printed actual newsletters – I learned to cut columns and paste them on boards with wax to send of pages to the printer, which is where the term cut and paste comes from – and made our money through direct subscriptions. We made enough money on each publication to support one or two writers, a shared office manager, and pay the rent in a pretty large office in SoHo in New York City. About half of the money went to printing and mailing.
At the time, getting the feeds and speeds and prices on anything was difficult, and those of us who worked at Technology News were very good collectors of information, which when used properly could save our subscribers thousands to millions of dollars as they configured systems and negotiated with vendors. This more than covered the cost of the publication.
The relentless pace of Moore’s Law and Dennard scaling meant that systems were getting more powerful and more capacious, often faster than commercial datacenters could consume each new generation. At some point in the 1990s, the price of storage and compute became cheap enough that the world could start storing all kinds of things – and it did. And then along came TCP/IP and HTML and the commercial Internet, and everything changed. Compute was cheap, and the value of data went down because things that were hard to gather became easier; information – providing insight – was still valuable, and if you gave enough of that insight, then you could still make a living.
You could make a very good living if you created a very good search engine for all of this information, as it turns out.
Making money with a subscription model on publications was pretty much done. The barrier to entry – those high printing and mailing costs – was destroyed by the Web, and so everyone had to switch to an advertising model to make money. The net effect of that over the course of a decade is that to make the same amount of money as we made in subscriptions required many orders of magnitude more readers and about an order of magnitude more content to have the numbers work out.
During the Dot Com Boom and in its aftermath, I wrote an enormous amount of stuff – somewhere between 1.5 million and 2 million words a year across the several publications I have worked for – and when the data analytics boom hit after the Great Recession, Nicole founded Datanami and was hammering the keyboard at about the same rate.
One of the reasons we founded The Next Platform together is because we wanted to go deeper into a story rather than to chase so many stories. We wanted to provide a higher value product, not a more voluminous one. Around the same time we were doing this, the value of data and the cost of storing it went pretty close to zero for people, but became of enormous value when aggregated across hundreds of million or billions of users. This is a trend that we both have independently projected would happen.
Here is a good, human illustration of how data is not worth much.
Do you know your home phone number when you were a kid? Of course you do. Mine is 201-293-3028. Who is the owner of the last phone number that I memorized, and that I will ever memorize for that matter? Nicole Hemsoth Prickett. What are the three numbers that I never memorized because I had a smartphone when they got smartphones? Those of my three oldest children: Ellie, Henry, and Chloe. I am pretty sure I won’t know Mia’s number when she gets a phone maybe five years from now. Here’s the point: I went from an oral tradition of memorization – of storing the data in my own head – to letting my phone keep track of all of this. You can call that progress, but something has been lost, too. My brain is out of the habit of storing data.
Search engines have destroyed how many funny and sometimes combative conversations around the dinner tables and bars of the world? We have been able to check any fact at any time for the past two and a half decades now, and how much human interaction did that eliminate?
Social media is another thing that is called progress, but maybe it cuts both ways. Social media does indeed allow us to keep in touch with each other as we have spread out across our nations in pursuit of work, home, and happiness. But the negative side effect of social media is that people don’t actually get together as much. We have gotten lazy about the physical thanks to the virtual, particularly given the larger distances we all live from each other once we leave the nest. And in some cases, people have been deepfaking themselves for years on Facebook and other social media to keep up appearances. We prefer actual interaction, although we must admit that the birthday reminders on Facebook are very useful for people outside of our immediate family and long-time close friends.
And now we are coming to the point where the value of information, the synthesis of data into a coherent story, and the value of creation are now a few points down further on a Moore’s Law curve thanks to generative AI. What set us to writing was the episode of the South Park cartoon that was released this week, all generated by AI Showrunner. No wonder the members of the Screen Actors Guild are on strike. All knowledge workers are concerned – from writers to lawyers to programmers – because these systems could evolve to replace them.
Here is a fun aside to ponder: Imagine how fast this would be happening if Moore’s Law was still going gangbusters, and how much faster it would be happening if money were free like it was only a couple of years ago. . . .
We think that the rise of the machines was always inevitable, but we also know it is hard to really feel an exponential curve. This is happening fast.
The people who have created these AI systems might be experts in complex mathematics, semantics, statistics, and system architecture, but that does not necessarily give them any higher authority about how this AI technology should be deployed or better insight than the rest of us about the effect AI systems will have on our work and home lives. In fact, the joy they get from pursuing artificial general intelligence, from the competition and collaboration with their peers in this high science, might blind them from thinking as carefully about consequences. We can’t tell you the number of times that we have talked to the executives at the upper echelons of the companies we write about here at The Next Platform and it always comes down to the same old arguments: We need to make money and if I don’t invent it, someone else will.
These are obviously true in every case, and the rest of us need to make a living – unless you want to go down some Black Mirror path where we are all on universal basic income consuming media all day. This is repugnant, and it smells of death and despair. Life is a gift, given to the living for such a short period of time that it is the most precious of commodities, as is this wonderful blue ball circling an average sun on one of the spinning arms of a perfectly normal spiral galaxy. It is so improbable that we are alive at all, really.
Remember: We are still a democratic republic, not a technocracy, and we all have a say in how we want our world to change. So, here is our advice: Think your own thoughts. Be tenacious about this, and be as smart as you can be. Your thoughts are precious and unique, invaluable as in beyond value, and that is why you were endowed by your Creator.
Note: Yes, the feature image is not an asymptote to zero. But I had never seen that one before, and it was more interresting.
Nvidia Proves The Enormous Potential For Generative AI
The exorbitant cost of GPU-accelerated systems for training and inference and latest to rush to find gold in mountains of corporate data are combining to exert tectonic forces on the datacenter landscape and push up a new Himalaya range – with Nvidia as its steepest and highest peak. It is …
Nvidia Research Gives Generative AI Images And 3D A Speed Boost
Time is money when it comes to generative AI, as is the case with most technologies. But with generative AI, it is big money. And the longer it takes for an AI model to do anything, such as training on data or generating images and video, the more money is …
With Blackwell GPUs, AI Gets Cheaper And Easier, Competing With Nvidia Gets Harder
If you want to take on Nvidia on its home turf of AI processing, then you had better bring more than your A game. You better bring your A++ game, several vaults of money, and a few bags of good luck. Maybe a genie in a bottle would help, too. …
Somewhere in-between Public Enemy’s “Don’t Believe the Hype”, and “Fight the Power”, one can probably find Rudyard Kipling’s “If” — to remind us to keep our heads when all about us are losing theirs to AI-hype, and so-aspire to replace Extras and cartoon shows with inexpensively uninspired synthetic automata, intangible plastic replicants, insipid concoctions of fartsy craftmachineship. Selling one’s physical appearance at a $200 lifetime value, to a “producer”, is a Fantine deal worthy of “Les Misérables”, a harbinger of Gavroches to come, waging “War at 33 1/3” for the cause of human dignity (I think).
Contemporary AI simultaneously promises to improve our lives and well-being, and to take-away our income and ability to decide for ourselves … what’s not to undislike on such a thoroughly well-planned road to a “Pretty Vacant” and dematerialized “No Future” of Sex Pistolism? “Efficiency and progress is ours once more …” as Jello (Biafra) suggested sarcastically … but here’s to hoping we’re not embarked on a 737 Max, and can redress trajectories for a proper mile-high club of superposed entanglement (or so I’m told)!
AI, like Indiana, and the “American Experience”, has to be what we make it … (granted though that it seems best suited for mass surveillance, and stock-market crystal-ball predictions, than anything else).
A heart-felt piece, and no doubt speaking to the fears of very many folks, not just those in creative spaces. While this AI megatrend is the cause of much angst, there is another existing and much longer running trend which has yet to really get on the radar like AI has. I’m speaking of the demographic transition and the coming (and likely irreversible) inversion of the population pyramid across not only the developed world but most of the developing world as well, all within the next couple of decades (where it hasn’t already arrived in force, of course). In the long-run view, it kinda seemed like humanity might be going to zero eventually anyway even without what we are currently calling AI.
Economy (or rather socio-economy) is essentially the business of things done by people, for people. The entire edifice of modern socio-economy is predicated on the extensive growth of populations. By extensive I mean increasing in absolute numbers, energy consumption, pop density, aggregate demand, money supply, whatever. That is about to come to an end, with really unknowable consequences for the viability of our extant economic systems. Intensive growth which speaks to system re-configuration, technology, doing more with less etc, isn’t the same, because intensive growth isn’t going to push more investment in new real estate, new energy production facility etc all of which sustain capitals structure which generally promises positive returns to investment. In fact, a declining labour force must increase the returns to labour versus capital, and I think we can see this trend asserting itself with the robust wage trend even while inflation runs high and investment is seeing generally negative real returns. The return to labour was already well out of balance in any case.
New AI capabilities potentially intersect with this trend as a way of mitigating the contraction of the global labour force even as it raises fears that it will increase the return to capital at the expense of labour (the latter being the fear expressed in this article). The only problem with assuming AI makes up for missing human workers is that normally, supply generates demand and demand generates supply, because it’s all people. Machines may generate supply but will not, until they become legal persons with their own unknowable goals, generate demand which is what leads to positive return on investment.
So when we imagine the future, we should strive to imagine a future which is defined by cloud-based AI, grey hair, constant de-population, climate change and a backdrop of generally negative returns to capital. Which may sound scary but it may not be as bleak for the average wage-earning meat sack as might otherwise be imagined. Of course it may still be bleak for writers but less so for, say, plumbers … but then, change in demand for different skill sets over time has always been a feature of history so what’s new?
Plumbers, electricians, and HVAC people are doing great where I live…. But we need to make money to pay them, right?
Well presumably the hvac guy and the plumber can buy one another’s services?
Thought-inducing ideas! I guess in France, we could require AIs to be unionized, which may increase the cost of their utilization, possibly beyond that of more normal labor (or maybe just tax it into the event-horizon of a nearby black-hole).
But the prospect of workers owning their personal means of production may be more palatable and sustainable — say, the AI that generates a likeness of you (eg. for a movie) would be owned by you (not by a movie producer), and you could rent-it out to one film production or another, following your own willful decision.
Or maybe even just the data that is that likeness of you, would be owned by you, and you’d be able to rent-it out for movies. We might also be able to rent-out our own personal web-browsing data (cookies, location, …), rather than just giving it away for free to the web data collectors and analytics outfits (it would have to be done willfully, as a contract, with appropriate compensation for each rental event).
From molten lead add grease plus classical lithography to A1, Lanier and Wang; self publishing, recording, videography but damn those direct mail costs still remain. Mike Bruzzone, Camp Marketing
With so many folks driven straitjacket insane by AI chatBot’s syllogistic tautological sophistry, nearing fullish ape- and bat-chit craziness and institutional certifiability (kind-of, ahem!, exaggerating a bit here) (eg. Lemoine/LaMDA, Weizenbaum’s secretary/ELIZA — appropriately coded in MAD-SLIP, …) — is conversational AI the LSD of the masses, by design? Might generative AI be the new cocaine of Wall Street (except for the “she don’t lie” part)? Does this town need an enema (again!)?
I’m glad that people with decades of experience in tech areas are expressing their concerns over what could become a bad trifurcation, instead of the good one which we all hope for!
@ Timothy,
You say,
“… some Black Mirror path where we are all on universal basic income …”
And I laugh a lot. UBI already exists, but is just hidden out of sight by,
1, Basic Tax Alowance.
2, Social Welfare.
And Europe knows all about it, and the actual “Social benifits” it brings.
Back in the 1980’s in the UK there was Form UB40 for “Unenployment Benifit” to be claimed. Back then quite a few people did, and contrary to what is claimed by certain conservate view points the were actually per head wildly more creative and productive than those in jobs.
You mention the “free money” for people to build with just a couple of years ago, in principle the idea was the same.
Except it was those of a certain conservative view trying to grab the benifit for themselves rather than society.
Their version failed as it always does for those with such conservative views because they fail to realise one of the basics of life. As humans on mass we “feed forward” and develop society such that it can grow. The modern version of those with a conservative view engendered mostly in the US is to coral both real –asset– wealth and fiscal wealth, for their own benifit, they mostly want to madly form society into some toy they control, but don’t have the mental capacity to understand their own failings.
Society which every one is dependent on is not formed by heroic strong men standing above the masses doing their bidding, you can see what that nonsense does over at the far east Europe where needless destruction on whimsy and greed is occuring. There is no profit in it, and no real social development and mankind will be globaly blighted by it for atleast the next half century.
What lifts mankind up is society interacting, where individuals trade with their skills and labour. Yes these change as society progresses but the achievement of all those little trades is society rises in oh so many ways. What has always held society back is that it’s been unproductive finding people to trade with. Taking your crops to market was a waste of productive time, not just for the seller but the buyer it’s why we just don’t do it that way any more. In the West these days most approching all trading is being done online. The reason shops still exist is why markets do, which is learning. People extend their reach in markets, in a way we just not have replicated on line, which we are still trying to learn.
LLM AI is not intelligence in any way shape or form Artificial or otherwise. Anyone who tells you it is, is either ill informed or lying to themselves or others. Yes there are the likes of Venture Capatalists trying to sell the next crypto-coin / blockchain dream but the reality is they are not selling the reality.
LLMs are a tool to do two things well,
1, Surveillance to extract information.
2, Imitate an organic indexing system.
Both are needed “to make markets for trading” in the way’s shops and old fashioned open air tablescand cloths on the ground still do.
Now you know what LLMs are realy good for, you can see how you can benifit society, and as part of the process not just you but those you “feed forward” to your family and social circle you trade with for that which is the mark of inteligence, “creativity and emotion”. All but a very few want to create, so they have a circle to trade it with. Those who don’t are by definition mentally aberrant, as with all things the aberrance falls on a scale which society imposes by observation on the behaviours of it’s members, and asigns a good or bad value, that changes as society changes. Generally what moves society forward is seen as good and what harms or holds society back bad. But it is also the observers position, their point of view that effects their individual view. Those of a liberal view are generally ebablers to the future, those of a conservative view are desperate to retard, hold back and corral society into worse than enslavement, serfdom of being property by their labours for nothing in return. As such those of hold back views are parasites, consuming without producing anything of real worth.
But as I said at the top I laugh, because you say accurately of you and your wife
” we are all on universal basic income consuming media all day.”
It’s what your job actualy is, the only bit you forget is the “creativity” you add from it, that you trade with others.
I feel like I consume technical material and produce media, for sure. But we actually don’t consume much media. Who the hell has time for that? HA!
I meant a universal basic income where people literally do nothing physical or mental (or both), and just stay at home watching TV or playing video games or watching the lives of others on social media, which I object to. Work gives life meaning, life gives work energy. I believe in that.