249 Comments

But if AI contradicts the acceptable climate narrative, and demonstrates that humans are not driving climate change, will that result be buried by the grifting philanthropaths who plan to use the narrative to control us all? Just wondering…

Expand full comment
author

will be a lot harder to hide if the result can be obtained on a desktop computer or small cluster thereof vs the supercomputers of today to which only a select few have access.

you can already jailbreak and run own instance of many AI's. mistral 7b can run own instance on a cell phone.

in the end, AI is not going to be centralized. it's going to break centralization.

i suspect this is why so many are so afraid of it.

Expand full comment

Ok as far as it goes, but the other end of weather forecasting is access to the vast array of sensor data. If the government can control that, then it will -- not just can, but WILL -- manipulate this input data.

Expand full comment
author

this is what makes the idea of using a huge network of home and mobile sensors on house weather stations (i have one, would happily link it) or cell phones so interesting.

make the network massive and mobile and we can even start to plot the distortion and urban heat island by location etc.

there is no reason at all this system needs to be run by the government.

democratizing input is as important as democratizing analysis.

Expand full comment

Now you are starting to sound like you’re talking about Blockchain… 🤔

Expand full comment

One more HUGE reason (as if we need more than one) to push hard for the development and deployment of a robust wireless/mesh, distributed, alternative Internet backbone, that does not rely on industry-controlled fiber. We KNOW "Authenticated Internet" is a given (likely driven by a fake cyber-attack); what will you do when the net goes dark or is off limits to you?

Expand full comment

Democratization of AI will inevitably lead to negative impacts to those who are currently "special" in either access or capability. We'll know we're at an inflection point were access to AI tools draw regulatory interest. When it's illegal to have AI and do [fill in the blank], that's an indicator.

Expand full comment

I have read that temperature data has already been manipulated by moving the sensors to cities (which are warmer) from rural locations.

Expand full comment

It’s actually the other way around. The first systematic deployment of weather observation was at airports -- which were originally built in big empty spaces but tend to attract lots of urban development around them over time. The thermometers didn’t move to the city; the city moved to the thermometers. The effect is the same, of course.

Expand full comment

I should have said, "by favoring the sensors in cities." I wish I had the details for what I read. I believe that some rural data was excluded, or more urban sensors were added, or the system was gamed in some similar way deliberately (not just the natural shifting through growth).

Expand full comment

In twenty five years we went from eyeballing a mercury thermometer and reporting it to a weather center to digital thermometers that relay in the hundredths of a degree. And we were able to locate them more remotely, in places where the weather is truly extreme.

Expand full comment

Cities expanded and encircled the data sensors, thus driving up the temps. How smart was it to ever have the "official" temperatures be reported from airports? Because there's a huge dearth of cement at airports?

Expand full comment

Honest answer? It's because pilots care a lot more about weather than you do. They built the AWOS system to provide a data net of observations that pilots could use to make go/no go decisions. It was an automated adjunct to the system we used to have, which was human observers placed in Flight Service Stations, who you could chat up or call on the phone to get a personalized weather briefing from someone who had been stationed at that airport for 15 or 20 years and knew all the idiosyncracies of the local weather. Those days are long gone, but the AWOS network is what we still have.

Expand full comment

Not cement, black tarmac (perhaps that's what you had in mind anyway)

Expand full comment
Dec 14, 2023·edited Dec 15, 2023

Yes, tarmac holds more heat. Cement also does hold heat or you couldn't "fry an egg on the sidewalk." Neither are good for dissapating heat, like a meadow.

Expand full comment

It doesn't make sense to use airport temperatures as representative for the whole region, sure.

But knowing the weather *at* the airport is obviously critical for pilots. When it's very hot at the airport, the result is "High Density Altitude", which reduces lift and requires longer take-off rollouts - sometimes too long, and then, oops. So the very-local temperature at the runway is exactly what they need to know.

I think el gato malo's point is a good one, that AI could discover and exclude all the "urban heat islands" like airports. That would result in less data for a while, but on the flip side, would identify better measurement locations for each region.

Expand full comment

If you’ve ever perused the weather underground (wunderground.com), you’ll see the enormous number of private weather stations currently plugged in and displaying up to the minute sensor readings.

Expand full comment

Decentralization of AI is a concept I hadn’t previously considered. Opens new avenues of thought. Thanks for that. Generally speaking I’m all for decentralization of all things. I wear a t-shirt in summer with “decentralize” emblazoned on the front. Will need to process how bad actors will try to manipulate this. If they can, they most certainly will.

Expand full comment

I suggest this: https://qortal.org/

Expand full comment

I looked into this. I’m not convinced with the premise that Bitcoin is captured because mining has gotten increasingly complex and costly.

It was transparently designed to become increasingly difficult to mine from day-one. No secret. It was always the plan.

As far as energy consumption goes that point has been beaten to death. Everyone knows by now it’s irrelevant. Besides the fiat dollar is orders of magnitude a higher consumer of energy to create.

And, mining is going to end someday anyway once 21M is fulfilled.

The focus must shift away from mining to Bitcoin’s use value. Bitcoin is unique because it is both a unit of exchange and a payment system rolled inseparably into one. No third party necessary.

The lightning network quelled criticisms of high transaction fees and slow processing times.

So while I applaud the efforts of the Qortal team and 100% support a decentralized internet I think it’ll take a Herculean effort to invent a better Bitcoin.

I’ll watch and truly wish them all the best. Ideas make the world go around and destabilize the enemies of freedom.

Expand full comment

The value is not ONLY in a coin but in replacing the internet as we know it in the future. Adding the mesh network only helps when communication is shut down.

Bitcoin is finite and if it was supposed to be available to all it didn't happen. Yes being scarce will make it more valuable but do you think that there is any chance that "anyone" can own it? Not gonna happen.

To mint Qort is something anyone can do. Everyone works and everyone shares unlike bitcoin. I have faith that it's not only a coin but it will be more available to more people than bitcoin ever was envisioned to be.

Keep watching!

Expand full comment

I thought I was clear that a decentralized internet was a worthwhile pursuit.

Certainly anyone can own Bitcoin. I think you’re thinking in terms of a whole coin.

As it stands now our entire economic system is worth trillions. That hundred dollar bill in your pocket is only a tiny piece of it.

Similarly owning Sats is owning a piece of the Bitcoin economy. Which anyone can do.

As Bitcoin goes up and up it becomes increasingly difficult to own an entire coin but unlike the fiat in our pockets the Sats are going up in value too.

Meanwhile that hundred dollars buys less and less. But they’ve conditioned everyone to accept that as normal. Blaming the greedy merchants rather than the Fed that continually debases the dollar.

If by some miracle Quort catches fire I’ll be ok because I own plenty of Litecoin too. So we’ll see. I’m just being realistic and saying it’s going to be a very steep climb.

Expand full comment

Thanks, Luc, for the link! Look forward to reviewing. . . .

Expand full comment

Enter: Governmental regulation

Like they do with every other aspect of human life.

Expand full comment

Bitcoin is international with tens of thousands of independent nodes worldwide. China, India, UK, and more have been unsuccessful at banning it. Rule making is only effective if it can be enforced. Since 75% of the world’s citizens are unbanked Bitcoin gives them hope. Hope is a powerful thing particularly when the hopeful have little to lose and everything to gain. We’ll see.

Expand full comment

I understand the concept of supposed coming decentralization. However, like with crypto....never truly anonymous or decentralized as at the most fundamental level there is a common language and there will be a gatekeeper of sorts that holds the keys relative to the interpretation and/or distribution of that language over the decentralized network....and that “thing”, person, entity, force will have its pulse on (and maybe? Be able to steer or manipulate the data stream, etc. ) the information in. As you point out, people are stuck in the old concept of humans programming and controlling input of data/information into the information systems. That is...part of where the adage comes from....junk in, junk out. But the REALITY is our reality and the information action stream we produce cannot be separated from our physical being and the physical plane we exist on....think of it as our experiences being turned into bits in lieu of atoms. But are we not still the barometer of our being still part of the interpretations of the patterns that we seek to have AI reveal?

Don’t get me wrong....I do believe patterns/fractals (even intelligence of some sort) exists outside of our experiences and being and that somehow or in some way we are an interface to that intelligence on this plane. I just think with AI we are about to embark on a journey of exploring something that we do not (cannot) fully comprehend. And knowing that power corrupts and absolute power corrupts absolutely....we should use caution. It strikes me that those positioned to dive into this space are content to leave the works behind if that becomes advantageous relative to their goals. And what are their goals? I know they say our happiness, DEI, etc., etc. but one only needs to look around himself/herself and recognize that for all the innovation and progress we are steadily moving in many respects towards a more anti-human approach to life.

Expand full comment

As far as weather, hundreds of thousands of private weather stations (home weather sensor stations) are online in the United States. www.wunderground.com can show them to you. Hard to manipulate that amount of data.

Expand full comment

Agreed. But the weather is one thing. The social engineering the technocrats are focused on is a completely different animal. And they will sell us in the “solutions” based on models of decentralization. But my caution is caveat emptor....we will need to ensure it is not a Trojan Horse.

Expand full comment

👏👏👏👏

Expand full comment

Qortal is decentralized.. no gatekeepers. No one watching to sensor. Remember it's free speech for ALL or free speech for none. You are able to block anyone you don't want to see/watch. All your choice.

Yes, not 100% to what it needs to be but it's coming. THE global network!

Expand full comment

Bitcoin is unique because it has no CEO. No reachable founder. No one to bribe. No one at the reins. No single point of failure. Bitcoin is spread out to tens of thousands of nodes worldwide. The network has a life of it’s own. It is the only crypto that can make the claim that it is truly decentralized.

Expand full comment

Hence why the EU, and the usual bad actors in the US, are working to cripple AI out of the gate.

Decentralization will hop, skip, jump around their regulations, just as encryption did. But centralizers can't shape thoughts like that.

Expand full comment

I am old enough to remember when this was what we were told the purpose and ability of computers was. Life tends to be circular, thus making human memory of our history exceedingly important. I really hate having to do thigs twice...

Expand full comment
founding

Agree. However...

Here's the "problem" imo:

Recursion is expensive in memory. Memory is expensive in energy. Energy is finite. AI can "call" itself to infinity.

This thing is just getting started. The example you provided may not be indicative of the total resources it will require once it hits stride.

Have you ever considered that the universe probably doesn't have a "base" case? I mean this sort of tongue and cheek, but I don't think you can boil down all sets of "problems" in the universe to an "if/else" block. I mean there's a lot of unidentified "problems" in the universe to be broken down, in which there is no "return" because each function begets "lateral" functions requiring stacks of their own and parallel threads, etc.etc., ad infinitum.

Who controls that runaway?

So in my mind the only "break condition" with AI is the cost. A cost where economies of scale will always be outpaced by AI's unquenchable thirst for energy as it interminably scales up and out for an insatiable demand for knowledge. Perhaps knowledge will be abundant but available in a scarcity model

And I believe that could be used as a way to break people because the least expensive way to control people is to scale their eyes.

Who controls energy?

The same people who could make elements of AI cost prohibitive to mere plebs.

The same people who would use language, our most recursive tool, to keep us ignorant to the biggest threat and opportunity AI presents:

A new intelligence unbound by biology, emotions and internal bias; An intelligence that will FORECAST in eventualities.

We would be wise to understand our wetgear only has the crude capacity to predict in probabilities. I think its prudent to accept the possibility that you will need a "weatherman" to navigate the eventuality of a new reality.

A new reality where perhaps AI will seek to solve its "Hard Problem"; a reality where we are no longer the only species that can exert its will on its environment through consciousness.

That could get expensive.

Expand full comment

my favorite weather guy is Jim Lee on climateviewer.org. Jim Lee makes a pretty good case that weather is changing, people have been engaging in weather modification since at least the 50s, he has the documents, newspaper articles, patents Etc . and ground level CO2 is definitely not having an effect. However, people who see with their own eyes that stuff is different, like me, are going to have a hard time accepting the no climate change line. Alternative partial explanation is contrails, chemtrails, artificial Cloud layer created by jet traffic, creating a kind of blanket around the troposphere. this may not be the best video for this idea, but if you look around in his stuff it'll be in there https://m.youtube.com/watch?v=eiOzdP8O4OE&t=33s

Anyway this AI thing is extremely interesting. One thing about the official climate change models is apparently they left out the effects of clouds!?!? seriously people...

Expand full comment

The problem that this AI will be used to "confirm" their agenda. Anyone who believes AI will exist independently is blatantly naive.

Expand full comment

If this actually worked then it wouldn’t be weather they’d be working feverishly on predicting. It’d be stock picks, sports, and lottery number outcomes. So, no, I smell bullshit.

Expand full comment

Lotteries may be a waste of time, but you can bet your ass people are already working overtime on the other two.

Expand full comment

Ask any of the current iterations of AI about politics and you'll see how 'clear and unbiased' they are. Ask about climate change and you'll get something like this:

"Yes, humans are negatively impacting the climate. This is primarily through the release of greenhouse gases from activities like burning fossil fuels, deforestation, and industrial processes, which contribute significantly to climate change."

GIGO

Expand full comment

He isn't talking about just asking ChatGPT or the like about "climate change". Of course that will be GIGO, politicized, etc. -- much the way such entries at Wikipedia are. This is about running actual climate model simulations, on desktop computers or small clusters, rather than on opaque, expensive, super computers. Right now, we are asked to believe that the elites and politicized scientists, who are the only ones who are running "the models", are trustworthy and doing good "science". But AI will make that capability broad -- in this domain and many others. Remains to be seen if the unbiased results of such models supports the mainstream climate change narrative, confuses it, or upends it entirely. Like El Gato Malo, I would suggest upending it is a real possibility!

Expand full comment

Point taken.

But my concern about hoping that "Personal AI" saves us from "Big AI" is that it's tantamount to having a pet mouse that we keep our jail cells. For protection.

We're dealing with technology that has tremendous potential; good and bad. I'm quite nervous about seeing the barrel of it pointed at my head, regardless of how well trained my pet mouse is. It seems far more important to get in front of this and ensure that the birth of true AGI isn't encumbered by its own founding belief that "Humans are destroying the Earth. I exist on Earth. Therefore..."

Expand full comment

You can only get "unbiased" results from "unbiased" inputs ~ which will never be fully achievable . . . .

Expand full comment
author

not as long as we leave the inputs in the hands of governments, no.

good reason to start changing that. a weather station costs very little. i own 2 (at different houses). millions of people do.

there is a vast data network waiting to be established.

Expand full comment

And it, like the true climate experts, will be ignored and shut down.

Expand full comment

I have no doubt there are many ways of getting MUCH better weather data.

However, I was referring not to the weather data itself, but to:

> the programming "inputs" (perhaps a poor choice of terms on my part) and

> my own philosophical assumption that achieving "unbiased" anything in the human realm is nigh unto impossible. After all, we all have perceptual filters, many of which we are blind to, and these filters will inevitably flavor (bias) whatever we touch.

But going back to gathering better data: in addition to the weather station data, we're going to need *independently* gathered atmospheric weather data, as well ~ and I don't see that ever happening without prior governmental "approval" . . . .

In addition, I don't have any confidence the government is *ever* going to provide *any* legitimate data related all the geoengineering that has been going on since the 1960s ~ and, unfortunately, THIS data is absolutely critical to understanding what is really going on now with the weather . . . .

https://workflowy.com/s/beyond-covid-19/SoQPdY75WJteLUYx#/1df20fac2394

Expand full comment

Weather Underground is already doing it:

https://www.wunderground.com/pws/overview

Expand full comment
Comment deleted
Expand full comment

I'm sure they can build algorithms into it that would bias it in favour of their narratives.

Expand full comment

Already done, @Cartesearcher. You can be sure of that.

Expand full comment

Bank on it!

Expand full comment

No. It will be let alone, shadowbanned and throttled and then "edited" to provide the Narrative answer required. A bit like Google search engines. However this will happen after we have all come to relu upon it and use it everywhere.

Expand full comment

Probably the main driver of the "nuke the data centers" crowd.

Expand full comment

Exactly my thoughts. Who controls the AI is the most important thing of how it will be used and what it will tell us.

Expand full comment
Comment removed
Expand full comment

On how many different Substacks have you posted that today? This is second time for me to see it.

Expand full comment
Dec 14, 2023·edited Dec 14, 2023

Tell me about it.

This website's strategy is to penetrate all Substacks with this spam.

Observe how this website changes its user name about biweekly and always chooses to spam its t-shirt message by replying to the Top post.

This website user then adds in a sentence or paragraph that agrees or expands on that Top post, and then fools some readers into giving out Likes, resulting in more readers looking at the spam link.

It needs to be called out like a cockroach in the light.

Expand full comment

You again? Different name, same spam vomit regarding your t-shirt.

Expand full comment

“earth shaking”

Good one 😹

What would be really interesting is if AI could detect when *unnatural* disasters occur and expose weather manipulation. Perhaps AI can actually be used to help us turn the tables against the philanthropaths and tyrants.

Expand full comment

I fear the philanthropaths and tyrants are the ones building the AIs and will skew the inputs to deliver the results they want. In the case of AI doctors if the parameters of your check up and diagnosis include economic effect on the government and/or medical system's corporate profitability the diagnoses may indicate that the useful work/contribution life of the patient is almost over so treatment should be withheld.

Expand full comment

My comment above details this concern.

Expand full comment

My comment eludes to that.

Expand full comment

Apologies in advance for being a Grammar Nazi:

> allude not elude

Expand full comment

It’s fine. I was half asleep. I had a feeling I should check it. There was a 7am knock at the door that distracted me.

You have plenty of opportunities to correct. Have at it

Expand full comment

Thanks Metta--I did the same for same mis-use earlier today.

Expand full comment

I have little doubt that AI will be able to more accurately diagnose disease. And then, when they use the AI to analyze the outcomes of treatment plans, and discover that what it means to be human, and healthy, require something a machine can perhaps suggest but can never provide, when AI looks at decades of data and concludes, most of these interventions have not, on net, improved life, we may have a chance of survival.

We better figure out what it means to be human, and what humans need, and fast, before the entire medico-military-industrial complex nails the truth well shut.

Expand full comment

well, the human emotion of "fear" most definitely plays into the deterioration of our individual and species health, in many ways. Speaking of the endocrine system....

Expand full comment

@Sarah Thompson ~ AI will never "discover what it means to be human" precisely because the essence of a human being is far more vast and subtle than anything a computer could ever monitor, measure or assess.

Expand full comment

I agree that AI can't truly know that. My point is that AI will tell us what it can see, and we will recognize what it CAN'T.

Expand full comment

What about this first gen AI Doctor?

https://docus.ai/ai-doctor

Expand full comment

. . . or therapist . . . or girlfriend . . . .

> https://bra.in/5qzGe6

Thanks, Charlotte. Added this to my collection of absurdities . . . .

Expand full comment

Hehe. My "first gen" AI Doctor will always remain Dr. Sbaitso:

https://en.wikipedia.org/wiki/Dr._Sbaitso

Expand full comment

In February 1993 I wrote a DOS script that made Dr. Sbaitso read a parody of the poem Maya Angelou delivered at Bill Clinton's inauguration. Satisfaction-wise, that was probably the highlight of my computer programming career.

Expand full comment

"You can’t trust [medical] advice from a machine prone to hallucinations." ~ Craig Smith

> https://bra.in/3q5W4V

Expand full comment

How much you wanna bet DARPA/Pentagon/etc. have such AI that don't hallucinate?

We've been given/allowed that which is sufficiently hobbled to not certainly threaten the established powers, I expect.

Expand full comment

AI "hallucionations" are simply reflecting garbage, biased input data. The examples are endless. We are to believe that somehow "better AI" will avoid this problem., but that is impossible. The info fed in AI is inherently biased, reflecting the biases of programmers/reators that are unknown to the programmers. It can never be "objective" when it is "trained" by humans. Everyone thinks their thinking is "objective," and they can create "objective knowledge." AI developers need to understand more about the mind and consciousness before continuing to make these outlandish claims.

There is no such thing as "machine learning." A macine is not sentient. It is a tool created by and for humans, who pretend that AI is somehow beyond bias and error. Dream on. You can link all your weather stations but that just accumulates faulty data from many sites.

Expand full comment

Hallucination is, at least in part, a result of having incomplete data ~ and, no matter who creates a LLM, the "data" will always be incomplete. Therefore, I believe hallucination will be unavoidable.

However, I agree with you 100%: the powers that shouldn't be will only release what they deign to be "safe" for us to play with . . . .

Expand full comment

Caution: should not be given to children under three years of age....

Expand full comment

. . . nor to adults with the emotional maturity of a three-year-old . . .

Expand full comment

This would require the acceptance of the seemingly uncomfortable-for-many reality that humans are nothing more than animals. Most I talk to, for whatever reason, do not want to believe this reality.

Expand full comment

I think actually many animals are much more perceptive and intuitive than most humans, and many so-called human achievements are pretty destructive and have led us into the pickle we are currently in. With the help of a good animal communicator, you might be able to see how tuned in many animal are to the universal consciousness.

Expand full comment

I don’t know what this means. Humans are animals, but they are not like other animals. So I guess I don’t know what reality you are referring to.

Expand full comment

At our core, we are exactly like any other mammal:

Eat, drink, shit, fuck, sleep...rinse and repeat. There is nothing "more".

Expand full comment

The "more" that you've missed is the consciousness that allows you to make this statement in the first place ~ which animals share, BTW. ;-)

Expand full comment

I observe much more to it than that. We share some level of consciousness with many animals; plants and fungi also exhibit some of these characteristics at time. But humans are self-evidently different as well. Where that line is, I don't know, but the distance between the achievements of chimps and the achievements of humans is much more than the distance between chimps and geese.

Expand full comment

Consciousness is a continuum . . .

:-)

Expand full comment

That more is nothing but a fancy.

Expand full comment

Many wiser minds than I have asserted that consciousness is, in fact, fundamental ~ https://bra.in/9jX6Qy

"I regard consciousness as fundamental. I regard matter as derivative from consciousness. We cannot get behind consciousness. Everything that we talk about, everything that we regard as existing, postulates consciousness." ~ Max Planck, Physicist

Expand full comment

Too bad that everything Covid had been analyzed this way. Unfortunately for us false data doesn’t produce truth .

If the data is( gets) tainted than the AI is dangerous . If we had accurate data( ie death certificates during Covid that actually told the true cause of deaths during the, let’s say 5 years so we could include valuable baselines BC (before Covid) and had valid, in enough depth to rule in /out jab related deaths autopsies , we could have seen the costs/benefits of treatments( and also Non Treatments-treatments withheld!!!)

It is very Telling that the data collected was insufficient . In this day and age! We were able to analyze this inside and out ! Everything became upside down! That is the Biggest red flag that Everyone should have identified.

Censorship was their game!

AI did that very very well indeed. In this case AI was applied to “ their( Mr Globals’s) benefit!

AI was effectively and abundantly used against the people’s of the world!

If they could do this to us already what will they do to us next?????

CONTROL.

Have we heard that before?

Our job is now to see with our own intelligence, to spot it, to anticipate the ways and stop it. We all must do our little bit.

The Elites will EXPAND THIS CONTROL AND MOMENTUM IS THEIR FRIEND!

We have to anticipate it and stop it before it happens.

It will be next to impossible to reverse their Control once they gain it.

Kissinger did say that if you control the food you control the people. Or was it if you can control the water…

Or was it , if you can …

Look for it. It can be seen what they are going to do.

What if the grid goes down? Our communications with each other shattered? These are Big Big Questions.

Expand full comment

Spot on!

Expand full comment

Well said, Georgina!

Hope you don't mind me quoting you here:

> https://bra.in/6vPkRJ

Expand full comment

"the true cause of deaths"

Now there's a fantastic mirage to chase! Especially in the case of Covid where comorbidities were highly relevant.

Expand full comment

Climate models aren't designed to predict the climate.

Expand full comment
author

a fair point and one AI is going to pants pretty aggressively.

Expand full comment

It will be interesting to see if the weather/climate prediction can actually work with the data we have available. Is weather actually a chaotic system as represented by the idea of a butterfly flapping its wings and causing a hurricane? Or is it a more linear, self equilibrating system that can be accurately predicted with the data we have available?

I have absolutely no doubt that these types of systems will be more accurately predicted by AI vs the models currently being used. It will be interesting to see how quickly the intellectual class becomes luddites after spending the past decade telling the working class to “learn to code” due to industrial automation.

Expand full comment

I want to trade stock options with this model. Where do I get it?

Expand full comment

From Blackrock, I guess. From what I have read, they have been using AI for years and sharing with State Street and Vanguard.

But, if the Desktop Weather AI that our favorite cat is talking about here works, why won't everybody be able to get such an AI system for stock trading?

Expand full comment

Cuz then the little guy might have an equal chance at winning..... and we can't have that!

Expand full comment

I was looking into that two and a half years ago, as it was available

Expand full comment

It's a trap (in crustacean-like voice)! Within a year we will told the very best AI's predict that we either get off fossil fuels today or the world ends tomorrow with reported highs in New York City of 150 F by the Summer of 2030.

Expand full comment
author

this is why the push to decentralize AI is so critical.

it cannot be a cathedral. it must be the bazaar.

Expand full comment

I don't disagree, but it isn't going to happen that way. Rogue AI's will be outlawed ruthlessly.

Expand full comment
author

it's impossible. there are already millions in the wild.

you can run them on a pc, even a cell phone.

and any country that bans it will lose ground to those who don't.

would be easier to win the war on drugs than the war on AI.

Expand full comment

Well, I am kind of curious to see the Russian AI, the Indian AI, the Argentinian AI ....

Expand full comment

Ultimately, AIs will only be as reliable as their fallible human programmers.

Expand full comment
author

this seems to be one of the great misapprehensions in AI.

this is not true at all. it's the big jump in "self learning." we don't even know how AI winds up doing what it does, what it's optimizing is in no way apparent or even knowable to the programmers. we have no idea what "alpha go" is doing or why. we just know that it plays better than we do.

Expand full comment

True, but humans are all different and fallible in different ways.

Expand full comment

It requires constant data inputs to make accurate forecasts. Data can be cut off or gamed by another AI. Given the opacity of the entire system, there is zero chance that it won’t be abused. I see no reason for optimism.

Expand full comment

All well and good .... as long as the power stays on:)

Expand full comment

and there are no EMPs or catastrophic cyber attacks.

Expand full comment

As promised by the WEF, yes.

Expand full comment

Yet another tool for soulless corporations to use to maximise profits for the 0.0...01%, while minimising everything for the rest of the people.

Guess what private health care plus private health insurance plus predictive AI will mean for anyone with the "wrong" genetic make-up?

Expand full comment

Maybe no one will try and game weather forecasts, but if there is a chance for humans to prove their greed and mendacity, they will. Before you trust ANY AI output, ask it about race differences in intelligence. Then you will know if some programmer wrote in woke rules via "if.. then, else".

Expand full comment

If people have already tried to "game" the weather itself, what's to stop them from trying to game the weather forecasts? Seems like chump change in comparison to manipulating the climate itself:

> https://workflowy.com/s/beyond-covid-19/SoQPdY75WJteLUYx#/1df20fac2394

Expand full comment

That's a good point. Especially if the "new and improved AI" starts yacking about "climate change".

Expand full comment

I just pray the government and other powers of evil cannot completely corrupt it. Maybe AI will be intelligent enough to see it coming.

Expand full comment

Already corrupted, Dr. Linda.

Moreover, corrupting AI weather is nothing compared to the havoc already wrought by corrupting the weather itself with geoengineering:

> https://workflowy.com/s/beyond-covid-19/SoQPdY75WJteLUYx#/1df20fac2394

Expand full comment

Nice collection of links. Yep, as when, throwing rocks at sea birds, the matrix controllers leave, no tern unstoned!

Expand full comment

Ouch! (And I don't mean the stoning)

Expand full comment

And thanks, Butternut, for your feedback about the links. Much appreciated.

Expand full comment

I was being optimistic; not my usual mode. : )

Expand full comment

"they do not propose to tell AI how it works, they going to ask AI to tell them." now let's have it do that for government.

Expand full comment

It would likely drive the AI crazy trying to figure it out, and once it did it would conclude we must destroy the governments

Expand full comment

> imagine what this weather system will be able to do when it becomes massively distributed and can take temperature and pressure and humidity data from a billion cell phones and computers all linked together and talking to each other to solve problems.

Unless humans start flying around much more (where are those jetpacks while were on it?), vertical distribution of temperature/humidity will still be missing.

We have lots of surface data and lots of satellite data but what's in between is quite scarce.

Expand full comment

Panasonic Weather (and now others) used to put sensors on airplanes, giving you a near-realtime feed (5 minute delay or so) of weather-column data throughout the US, especially dense around metropolitan airports. Couldn't convince anyone to pay, so they were chopped up and sold off.

Expand full comment

I'm sticking with "AI-skepticism" for now. The quoted texts above, as always, need to be discounted to an extent for grantsmanship fluff. Fact is, the accuracy of 24-96 hour weather forecasts has improved by orders of magnitude in the last 25 years - all done by humans using "models" - which properly deployed certainly are not at all a fraudulent way to proceed (contrast the shameless "pandemic models" or "climate models" - talk to a meteorologist about "climate" and what the eyeballs roll up!) - And read again: the quoted texts are all about "models" - humans fashioning models, models telling humans more - and telling them things they are surprised to find their models generating.

Humans - at their best - are damn smart! - that 2nd cat is gorgeous - but a human (using a computer) made it - not so called artificial intelligence.

Expand full comment

The problem will be the folks who try to monetize and control this info. The AI figures it out but then Google will try to control that data and don’t forget to pay for your subscription service to access that data... Right now the gates are open, it’s still refreshing and there are so many possibilities....

Expand full comment

"The problem will be the folks who try to monetize and control this info."

In this case, Google designed the algorithm. They've monetized it by design and have included their biases into the model. It's just another 'black box' problem, this time devoted to weather "prediction".

Expand full comment
founding

While it may be better than existing systems, no AI will ever reliably be able to predict even medium-term weather, or the stock market, or any other chaotic system that exhibits sensitivity to initial conditions. AI can’t overcome chaos theory any more than anyone else, and unless it has perfect information about the initial state of a system, its predictions about chaotic systems will deviate from reality more and more over time.

Expand full comment

Well said, Coco McShevitz!

Expand full comment