Welcome to the cranky uncle’s website & microblog!

Demand  Imagine you were a trader.  You set shop in a rural village, and tell everybody: “I'll buy chickens for $8”.  You plan on reselling them for a profit in the big city.

Farmers will make a quick calculation: “I can raise chickens for $5, so I’m bound to make a profit”.  Thus you have created a demand for chickens; farmers will raise them for as long as you buy them, and as long as you buy them, they will produce more and more chickens.  You sell them in the big city, everybody happy.

Of course, this works negatively as well: remember that (probably apocryphal) story about the British in India, who paid rewards for every cobra captured, so Indians started raising cobras.  Thus the “perverse incentive” is created.

Incentives quickly become disincentives when money is involved.  We all want to help single mothers in need.  No one wants children to go hungry.  But generous welfare soon becomes a demand for unwed mothers and fatherless children, if a “welfare queen” could improve her standard of living by having more fatherless children.  Knowing of the detrimental effects of the absence of a father in children’s lives, society doesn't want that!

So society must be very careful in creating subsidies, because not only we could be creating perverse incentives, but also dis-incentivizing positive behaviors.

As I've said many times, poverty doesn't create rights!  Every society has always tried to provide for the extremely poor and the homeless.  These also had an incentive to be agreeable and not create problems for their hosts.  But when welfare becomes a “right”, the incentive to be agreeable to your patrons, disappears.  That's why subsidies are so dangerous: people quickly get used to them and feel entitled to them, and if the mere idea of phasing them out is entertained, those on the receiving end get immediately incensed and even justify violence.

I have also the suspicion that a generous welfare state is the main disincentive for having children.  You don't need children to look after you, if you have a reasonable expectation that the state will take care of you.  The welfare state is also a draw for the wrong kind of immigrant; a country can't have both, a generous welfare state and open borders.  But that was the Lefties' plan all along: destroy society, including the welfare state.  Remember, “the issue is never the issue; the issue is always the revolution”!

So, beware of perverse incentives and entitlement!  Most societal charity should be private, not public!

Damned phones  Computing should be done in the computer!  Just as the TV promoted mindless consumption —as opposed as reading, which had you imagine worlds, and reading news allowed you to reflect on them, instead of quickly cutting to the next, as on TV—, the phone promotes mindless scrolling and swiping.  You can't create on a phone.  You can't write on it.  It dissuades you from expressing your ideas.  But it has some magic incantation that draws us so powerfully to it!  It must be actively resisted.

I've used tablets for over a decade now.  Phone screens now look uncomfortably cramped to me.  But they look so alluring to me anyway.  Reading on a tablet is akin as reading on a book, and watching movies in it is feasible as well.  They are clunky and can't be carried in a pocket, though; maybe that's for the better!  You can't just mindlessly pull out a tablet to fill a few seconds of idleness.

I understand the need for some to be constantly connected and on the go.  Busy executives much?  But if that's not your case, I encourage you to switch to a tablet.  Anyways, most of the computing should be done in a laptop or desktop PC.  Keyboards encourage you to express your ideas more fully.  

You don't need to go cold turkey and ditch the phone (but you could totally, if you wished).  Make a conscious decision to use the laptop more!  Use the web instead of the apps!

The push to watch Youtube on TVs also dissuade you from interacting.  You aren't going to comment, or copy a link, using your remote control.

Not passing the baton  Many youth all over the world complain that home ownership is becoming less affordable and jobs are harder to come by.  What would be the cause?

I contend that the generational shift isn't happening correctly.  Boomers should be transferring their wealth to millennials (us Genx'ers are used to being forgotten).  Search for any graph and you'll see.

Once I read that «a homeowner doesn't riot», and that rang true!  It's only the descamisados with nothing to lose, that destroy their own communities.

It's also observed in many cities that there are many abandoned buildings and homes.  A lot of people have migrated and some even are decades in abandon, while many young families struggle to buy homes.  So what should be done?

First, allow the quick transfer of real estate from “dead hands” to young, married couples with children.  Foreclose properties with unpaid taxes and auction them to young married couples with children.

Second, relax zoning laws.  I've heard from many sources that one of the advantages of Tokyo, is its relaxed zoning laws: you can build a workshop, a store, a school, a house, near one another; so you can live and work nearby.  There is no single “downtown Tokyo”; there are many different zones, all with the services their inhabitants think they need.  Also, relax building codes; this will also allow people to build the houses they can afford for cheap, instead of the “optimal houses” that bureaucrats dream from their ivory towers, but are unaffordable to real people; and it will discourage invasions.

Third, ease entrepreneurship.  Though I haven't come across to the data yet, I'm afraid millennials and zoomers aren't starting enough companies.  They were encouraged to be employees, but now there aren't enough companies to work in!  Are young people building the next Amazon, the next Walmart?  I don't think so.  So we must relax laws so those new companies be created.  Flat income tax 10% (I think it was Phil Laut who said, “if 10% is enough for God in tithing, it must be good enough for the government!"); no minimum wage laws so jobs can be created; 10% VAT and that's it.  This, coupled with the relaxed zoning laws we mentioned, must help create new businesses as needed.

All this can be summarized in: we live in a world of rules set by boomers, for the benefit of boomers; but these rules probably won't work today for millennials and zoomers, with the challenges they face.  So Millennials should try the rules that fit them.

No humans  An article mentions another AI flaw, which is its lack of human flaws: LLMs have no skin in the game, and they are infinitely patient, which almost no human is: “Attention without scarcity isn’t a trust signal. It’s suspicious in the same way a pharmacist who hands out unlimited pills for free is suspicious”.

Want another paragraph? Here you go. Another page? You’re welcome. Another two million words? My pleasure. It seemingly doesn’t care about anything — which is exactly right — and thus feels hollow.
Finally, there’s yet another reason — we care about the human touch. Think of it like calligraphy vs a computer font. You can generate a thousand pages in a thousand different fonts with a few clicks. It’s efficient. It’s clean. And it leaves no trace of a hand. Calligraphy is messier. It’s slower. It has tremors, pressure changes, hesitations. But that’s exactly why it works as proof of a human being there.

In the end, it’s actually pretty simple. Humans trust humans. Humans do not trust soulless agents. That’s why calling someone a robot, automaton, or machine is rarely a compliment.

AI is also the master of ersatz:

You know how you can tell the difference between a photo of someone genuinely laughing, and a photo of someone saying “cheese” for the camera?  One has visible joy. One has visible effort to look joyful.
AI is really good at producing “say cheese” writing. It can nail the surface. But it can’t produce the equivalent of genuine laughter, because genuine laughter requires a real person actually finding something funny.

I guess writing has to be, more than ever, original.  If AI can beat human writing in “polish”, so to speak, we must become even more human.  AI can’t have experiences (at least not yet!), so that’s what should make our writing valuable.

Oh how quaint look the traditional rules of yore: the Turing test, Asimov’s Three Laws of Robotics in the era of AI murder drones.  But this author gives the Turing test another twist:

But here’s the thing about AI: it can fool you for five minutes. Maybe ten. Maybe even an hour. But talk to a chatbot for a year and there’s no way you’re going to believe it’s a human.  Yes, AI is very good at producing a convincing piece of writing. But it’s not good at producing a convincing writer across texts. In blogger terms, AI is great at sounding human in one post. But it struggles sounding human across a body of work.  And the reason is simply that large language models are consensus machines.
And the reason is simply that large language models are consensus machines. They look at immense datasets of text, draw a big circle over a certain area (e.g. self-helpy stuff, personal essays, opinion pieces), and then remix everything within the circle to produce a self-help piece, a personal essay, an opinion piece, whatever — without ever producing someone’s piece.
They do this by basically throwing dice: say this thing with 10% probability, say that thing with 5% probability, have this weird writing tick with 1% probability. But human writers don’t throw dice each time they sit down to write. They just show up as themselves, every time, whether they like it or not, and don’t use probabilities to decide whether today they’re going to wear a Che Guevara t-shirt or golf attire; or whether they’re going to be a single mother with three children or one of the children in their 30s still using the family Wi-Fi to lurk on 4chan.
Put simply, ask an AI to write 200 posts and it’ll produce 200 posts. Ask a human to write 200 posts and you’ll get a person slowly revealing themselves whether they meant to or not.

There’s the theme again: by definition, AI has no personality; it has no experiences, no flaws, no human limitations, no cramps, no aching bones, can’t reflect on the inevitability of death etc.

AI can be used as a tool, an editor, not as a replacement of the human behind the pen:

if you’re using AI as a critical editor after you’ve already written the piece, or if you use AI as a tool to interview you, well, I think then you’re just doing the same thing professional writers with resources have always done.  You’re doing what Stephen King and Margaret Atwood did when they (or their publisher) hired an editor. Or you’re doing what J.R.R. Tolkien did when he regularly sat down in a pub with C.S. Lewis to talk his ideas over.  AI just lets you do the same when you’re broke and have no friends.

Humans crave human experiences.  That’s why we follow writers, artists, celebrities:

Computers crushed chess decades ago. And yet… Nobody watches computers play chess against each other. Not because it’s bad. But because nobody cares what a perfect decision looks like when there’s no one inside the decision … So the point is not to beat the machine. The point is to still have humans playing in a way that only humans would play. Same with writing: writing is no longer about producing the best possible text. It’s about producing text that clearly comes from one specific human mind over time.  And this is nothing new.  Not only do humans watch human chess even though computers are better. We also watch sprint races even though motorcycles are faster. We watch strongman competitions even though hydraulic machines exist. We watch people compete in hot dog eating contests even though industrial food processors exist.  They’re there for the human effort. That’s what fascinates us.

In these uncharted waters, might it be useful to write using software that can trace the writing process, as evidence of human work?  Will it be necessary to write in Google Docs from now on?  Will we have to abandon our old but trusty text editors?

Silly times we live in.

There’s no point anymore  Up to now there was some hope that the public would appreciate real effort in creating art.  But an article made me realize that «You can’t give a chimp a machine gun and expect it to never pull the trigger. And once it does, that power trip will permanently sedate any conscience it might’ve had before … Also, most people aren’t really people anymore in case you haven’t noticed. They’ve already merged themselves with the machine. They love the AI slop, and can barely tell the difference. How many people do you know with an actual taste for good writing?».

Those who read are in a minority; and those who are able to discern whether words are human or AI slop, are fewer.  The unwashed masses watch fruit dramas and enjoy them!

AI was a Pandora box begging to be opened.  There never was a possibility of not letting the genie out of the bottle; our curiosity always gets the best of us.  AI has lifted the veil of many absurd tasks, though.  The fact that kids overwhelmingly use it for homework, gives you an idea of how pointless and repetitive it is.  “If you want to see the future, look at the young”; they’ve embraced AI with gusto and not scruples.

Even talented and smart young girls have embraced AI as their psychologist.  They do math homework by taking pictures of the exercise and asking AI to solve the problem.  They do English homework in the same manner.  The challenge now is finding tasks that are meaningful.  Human interaction must be meaningful from now on!

Heck, that quote three paragraphs above could have been made with AI, and I wouldn’t have noticed!  Which takes us to the next idea…

Another good writer makes a very compelling case about the futility of AI detection software.  I used to think that it was a big “Gotcha!”, but the reality is that most people couldn’t care less:

. The share of people who give a shit about whether a given piece of writing, art, or code was produced with AI is going to fall off a cliff, and that fall has already started.  Five years from now, asking “Is this AI?” will feel about as urgent and meaningful as asking whether a photo was taken on film or digital. A few specialists, myself included, will care intensely. Everyone else will have moved on.

I guess it’s due to an increasing demand of mental energy, to determine whether words are human or AI-made.  Most people won’t be bothered, and will only care about the usefulness:

The “is this AI” question was never the question; it was a shortcut around better questions [“Is it well written?  Is it useful info?]   When the correlation breaks and the proxy decouples, people will abandon it. It has simply stopped being a useful measurement, and useful measurements get dropped without ceremony.
None of the millions of people who buy James Patterson novels actually care that James Patterson hasn’t written a James Patterson novel for the best part of a decade; he simply hands outlines and ideas to a ghost writer who produces the finished work …
For the vast majority of people, and the overwhelming majority of the writing they actually consume, the manufacturing process is not actually a property they care about for its own sake. An airport thriller, a recipe, an assembly manual, a product description, a weather summary, etc. — the question “was this written by x human” is already uninteresting. It was uninteresting 12 months ago, and it’s uninteresting now. The questions that matter to us are simple — does this content communicate something useful / valuable / reliable? Is this content enjoyable?  The author was rarely the point; why else would people complain at length about recipe creators injecting their personal narratives onto the page?
Moral panics are largely temporary. Almost everything we were sure would rot our culture, our brands, and the world around us became background noise within a generation … You can’t sustain a moral panic about something that feels like it has just always been there. For people turning 16 from now to the next decade (and I regret to tell you, there are more of ’em every single day), AI will have functionally “just always been there.”
If you’re sweating a pangram score [AI detection software], I think you’re optimizing for a thing that is about to stop mattering at all; learning to fool a metal detector that is just going to be unplugged anyway.
If a blog post meaningfully changes your life, gives you the strength to get sober, inspires you to create — I don’t think you’ll care about the provenance as much as what you gained from reading it. And what you gained is entirely up to you to measure and define.

I guess purism isn’t realistic anymore.  The tools are there and if one refuses to use them, one quickly becomes outdated.  Employers expect you to use them, and are exasperated when you’re not aware of some AI tool or workflow.  AI-Luddism is not appreciated.

Condition yourself  An article [V.] suggests “conditioning yourself” like Pavlov’s dog in order to “get into the writing flow” more easily.  He says he got used to listening to Scottish bagpipe music, and now as soon as he starts playing it, his mind sets in “writer mode”.

He adds that he doesn’t particularly enjoy that kind of music in particular anymore, but after the conditioning he has subjected himself to, it’s very helpful for him to start writing.

He adds that it probably has to do with the fact that the brain prefers the path of least resistance, and since “repetition solves a problem.  The brain spends all day making decisions, so whenever it finds a sequence that seems to produce a useful result, it often keeps the entire package together. Write enough articles while drinking the same coffee and suddenly the coffee becomes part of the process. Repeat the pattern for long enough, and eventually, the words flow from my fingers like a pianist reciting a piece of music”.

He finishes with “How an article performs is outside my control. But showing up isn’t. And neither is finishing. So, when I look back over the last seven years, that’s what stands out. The fact that thousands of pages exist today that didn’t exist before. And, surprisingly, many of them were written to the sound of bagpipe bops”.  Well put!

A mere novelty  As usual, Doctorow [V.] has very good insights about AI:

all AI really has is potential. Actual, existing AI is useful at the margins, if you're already a skilled practitioner who can discern correct from incorrect outputs, and if you integrate AI judiciously so that it doesn't overwhelm your ability to pay attention to those outputs and apply your discernment to them.

Reminds me of the supposed “potential” held by 3D printers and crypto a few years ago.  Of course so much potential failed to materialize.

In other words, AI is mostly a novelty, a heavily subsidized toy that produces little more than distraction. Where AI does produce value, that value is comparable to a plug-in, a new feature for your word processor or image/sound/video-editing package that might help you do your job somewhat better, or it might not … That doesn't make AI useless, it just makes it a normal technology: useful for some, useless for others, capable of being abused and likely to waste a lot of time when used unwisely.

There!  A normal, “meh” technology that one can find a few uses and period.  Not something that’s going to change the world entirely!  A normal technology that normal people will use to lazily avoid doing homework, to make terrible flyers and memes, and not much else.

The “plugin” metaphor sounds useful.  If you already know what you want (and even better, if you already know how to do it), AI will enhance your talents, and help you get past tedious tasks to achieve your results.  I’d say it will show in the results: if AI is an enhancer of present ability, or an enabler of mediocrity.

Just as 3D printing and crypto: they didn’t revolutionize the world; a few niches were found where they were useful, and there was that.  The rest was hype and bubble.

Above all, we need to stop talking about AI as exceptional. AI is normal. A normal technology has some uses, but isn't useful for all things and all people. A normal technology isn't inevitable, it's something you decide whether you want to use or not.

Exactly!  So far it’s a slop making machine.  Undoubtedly skilled programmers will find it useful (I myself, not being a programmer, have), but for the unwashed masses it’s a distraction and a net negative.

Finally, I couldn’t best his characterization of that reptile Altman.  Seriously, go read the article, it’s that good.

Seven rules  Another article summarized Vonnegut’s rules for good writing: The writer’s «first job is to grab the reader’s attention and their second is to take care of it so they don’t leave».

The rules: 1) Use people’s time in a way that makes them feel it was time well spent; “words that must be better than silence”, as I say.  A lesson, an emotion, an idea, a new perspective, all are valid.  2) Give the reader at least one character to root for.  Then they can empathize and invest emotionally.  3) Every character should want something, for desire is the heartbeat that keeps a story alive.  4) Every sentence should either reveal character or advance the action.  5. Start as close to the end as possible.  «Starting in the middle of the action … is exciting for you as the writer and transfers that same pull to the person reading».  6) Be a sadist.  No matter how sweet and innocent your characters are, make terrible things happen to them.  7) Write to please just one person.  Think about your ideal reader, write only for them, and let the others find their way to you.