So what's next after AI?

Maturin said:
I can only partially agree. Depends on the reader. Depends on the fiction.
A good science fiction in my eyes does not implement funny ideas in peoples heads, but is a reflection, a criticism of a current situation
Click to expand...
Yes, of cause it always depends. But for acquiring capabilities for being able to criticize in a constructive manner one does not need to consume science fiction.
I prefer to select carefully what I allow to enter my brain.
 
Criosphinx said:
Still far away from answering The Last Question
Click to expand...
If the "last question" was simply:"what comes beyond the universe?" it's a good example of physics vs. science fiction.
sf can deliver thrilling, gripping, entertaining answers, while physics come with boring real truth:
"This question cannot be answered, because it's outside of this universe we're in. By principle only questions inside this universe can be answered."

POSIX.1 said:
I prefer to select carefully what I allow to enter my brain.
Click to expand...
Cannot agree again. Especially when I think back when I was young, what I all allowed to enter my brains... 😂
Personally I'm neither a big fan of science fiction nor phantasy. There is way more books than you can read in a lifetime without picking any kind of that. But besides it's entertaining - one cannot read only technical literature, biographies, self coaching, politics, philosophy or documentations, only; a good mixture is important - good entertainment literature is also always some kind of philosophy.
But I agree insofar: If some one reads science fiction, only, has no reference to reality, this was stupid, indeed. But I see similar signs with phantasy readers, but also hardcore computer game players, or people abusing massively drugs... This are all kinds of fleeing reality, or losing contact to it unnoticed.
But that's not the issue of neither the books, nor the games, nor the drugs...
And I personally, especially when reflecting my own youth, prefer 'abusing' books above drugs. 🥸
 
Switching back to the OP's original question... interesting list here of the top 5 hacker predictions for 2025:-

hadrian.io

Top 5 Hacker Predictions for 2025

Discover the top 5 hacker predictions for 2025, including AI-powered exploits, IoT threats, phishing tactics, and Web3 vulnerabilities. Stay ahead of cyber risks.
hadrian.io hadrian.io

Let's hope running freebsd makes me less vulnerable to some of these!
 
And Uranium has been touted as one of the next big things for the last year or so. This is based on the notion that the energy consumption of data centres is projected to grow exponentially if AI gets widespread adoption. New investments in nuclear are being at least considered in europe and the US, with companies like Microsoft considering investments in data-centre localised SMR's. If any of this comes to pass, all of the new reactor fleet will need uranium... hence uranium miners and refiners should do well. We are talking about companies like BHP, Cameo, Denson Mines, etc.

www.cruxinvestor.com

Uranium, Uranium, Uranium: The Strategic Investment for 2025 & Beyond - Article | Crux Investor

Discover why uranium is the ultimate clean energy investment for 2025. Explore market trends, geopolitical impacts, and standout companies like Energy Fuels, Global Atomic, IsoEnergy, ATHA Energy and Premier American Uranium.
www.cruxinvestor.com www.cruxinvestor.com
 
Chip manufacturing is another area ripe for development. Right now, a handful of companies dominate, of which Applied Materials, ASML and TSMC hold the global crown jewels. This is a market that is ripe for new entrants to challenge the dominant position of the incumbents. SMIC and Huawei come to mind, but I'm sure there are other companies wanting to take some of that business, including Intel.
 
blackbird9 said:
And Uranium has been touted as one of the next big things for the last year or so. This is based on the notion that the energy consumption of data centres is projected to grow exponentially if AI gets widespread adoption. New investments in nuclear are being at least considered in europe and the US, with companies like Microsoft considering investments in data-centre localised SMR's. If any of this comes to pass, all of the new reactor fleet will need uranium... hence uranium miners and refiners should do well. We are talking about companies like BHP, Cameo, Denson Mines, etc.

www.cruxinvestor.com

Uranium, Uranium, Uranium: The Strategic Investment for 2025 & Beyond - Article | Crux Investor

Discover why uranium is the ultimate clean energy investment for 2025. Explore market trends, geopolitical impacts, and standout companies like Energy Fuels, Global Atomic, IsoEnergy, ATHA Energy and Premier American Uranium.
www.cruxinvestor.com www.cruxinvestor.com
Click to expand...
 
Hahaha, yeah, I've read loads of articles touting uranium over the last 6 months 😁

And this is why DeepSeek was a big shock to the market early this year; they appeared to be able to get the same AI results without having to use ultra-power-hungry GPUs. But after the initial announcements further stories came out saying they had used nvidia's top chips after all... so who knows what the real position is (other than the guys who work for deepseek). Of course if you can do it all on standard chips then you don't need a nuke next to your data centre to provide a GW of electricity...

Rare earth miners are another fever stock. Deep sea mining of sea-bed nodules off the coast of Japan... opening new rare earth mines in the US... rare earth mineral recyclers... etc
 
Speaking of nuclear, I wished fusion would finally have that big breaktrough in the near future. Maybe we could use AI to develop a proper, easier way to make that possible.
 
There have been lots of announcements on fusion over the last couple of years, for example.

en.as.com

Canada turns the tables on Trump: breaks all records in nuclear fusion, sparking unprecedented scientific optimism

This groundbreaking experiment employs cutting-edge cost-effective technology, which represents a step closer to achieving clean and affordable energy.
en.as.com en.as.com

As Sabine Hossenfender pointed out, none of them are anywhere near to sustained net energy production at present. But people keep plugging away... the prize is astronomical for whoever manges to crack it.

Here's Sabine's latest message on fusion. She likes stellerators :-)

View: https://www.youtube.com/watch?v=e7hw0aC1BbI
 
And solid state batteries are another 'real-soon now' next big thing. CATL, BYD, Toyota, Samsung, LG Chem... all working on this. The claims are charge in 5 minutes and then 1000 miles range in your EV (well, figuratively speaking). This is one that I hope happens in order to make EV's start to be a realistic competitor to FF. Especially if the damn things don't catch fire with the alacrity of the current generation of batteries. If this works out, it might also be a key enabler for grid-scale (non fossil-fuel) storage, which would be another game-changer.

CATL bet on solid-state becomes reality – Batteries International

www.batteriesinternational.com www.batteriesinternational.com
 
POSIX.1 said:
Yes, of cause it always depends. But for acquiring capabilities for being able to criticize in a constructive manner one does not need to consume science fiction.
I prefer to select carefully what I allow to enter my brain.
Click to expand...
Totally off topic...

Interesting read about Grothendieck...
www.theguardian.com

‘He was in mystic delirium’: was this hermit mathematician a forgotten genius whose ideas could transform AI – or a lonely madman?

In isolation, Alexander Grothendieck seemed to have lost touch with reality, but some say his metaphysical theories could contain wonders
www.theguardian.com www.theguardian.com

And Edward Teller talking about John Von Neumann. I always thought Ed Teller was a great speaker.
View: https://www.youtube.com/watch?v=Oh31I1F2vds
 
Criosphinx said:
Still far away from answering The Last Question
Click to expand...
I used to think the answer to Asimov's last question is life itself, in that life is the most complex form of organisation of matter that we know of, and life evolves into more complex forms over time, so I naively thought of life as a kind of negative entropy. Consider for example the persistance of the DNA molecule and the patterns it encodes over geological time; let alone the complexity of a single cell. But of course in reality the entropy in thermodynamic terms of the whole system increases inexorably and the high complexity of living systems is superimposed on top of the underlying physical system.

Drew Berry in Melbourne has done some great visualistion animations of cellular processes.

View: https://www.youtube.com/watch?v=7Hk9jct2ozY

And the famous flagellar motor

View: https://www.youtube.com/watch?v=dYt5135_0bs

How do raw atoms and molecules become arranged into these structures? And this is just the substrate from which our own consciousness apparently arises? It's certainly beyond my comprehension.
 
SirDice said:
Speaking of nuclear, I wished fusion would finally have that big breaktrough in the near future. Maybe we could use AI to develop a proper, easier way to make that possible.
Click to expand...
Sorry, to kill your illusion, but that's unrealistic. That's kind of a physic's playground, but nothing society can rely on as an actual solution within a realistic future.
We're told that again and again, because a) physicists need lots of (tax) money for their 'toys' and b) it's one of the resorts those withdraw to who simply cannot grasp nuclear energy is an artifact from last century's cold war. The original idea of nuclear power plants was to mass produce nuclear weapon's material. Electricity is a by-product. But the work made to bring democratic societies to join this in the 1950s and 1960s was so convincing the idea is still hard to be erased out of some heads again.
However:
If science would make the crucial breakthrough that is needed to build fusion power plants that are practically usable (efficiency, economic aspects) - and we are by far not there yet - so the physicist have done their job, now the engineers take over - you still have to accomplish at least two things:
1.) design a powerplant that's suitable as a real power plant for a real market, get it approved, plus get the money for it (app. at least ~10 years)
2.) build it (again at least ~10 years)
So if there was this needed breakthrough tomorrow, there will be at least twenty years (and this estimation is rather optimistic than realistic) before the first electric power is fed into the net.
Don't get me wrong: I don't say 'don't do it at all' I only say 'it's no option to solve any problem of our current situation in time.'

Fusion energy as a realistic solution is a dream for those who will not accept the fact we need to reduce our energy consumption. Looking for and betting on energy sources alone will not work.
Point is: Our current energy consumption still rises exponentially unbroken, and undamped. Not only because of cloud computing, AI, and data centers, but also because of electrical cars. [yeah, shit, right?!] but mostly because of our consume still rises.
Fact is there is no energy source we can expand that fast as our energy consumption rises at the moment - no solar, no wind, especially not nuclear (big plants, yes, but slowest in building watt per time; too slow (even those 'small modular series production' plants that always come at this point are no help) face reality: nuclear power is riding a dead horse). Not even a chance to keep up if we still rise our energy consumption as we still do. And we do. The only energy source that can keep up is fossil energy sources, only. Because there already is enough infrastructure to exploit the resources, and oil, coal and gas plants can be built quickly enough in time to keep up. (It's somehow logical, becasue it was the fossil energies that enabled that fast exponential growth we experienced, and it's so hard to get rid of it.)

Don't get me wrong!
I am not pro-fossil energy. But I am realistic enough to look at the facts and see: Without reducing energy consumption there is no way, not a chance to get out of this trap.
Or to put it briefly:
Drive smaller cars, not even bigger ones and then hope for some miracle (like fusion energy)!

To get back to topic - and in this case of nuclear power I join USerID's idea of an 'evil masterplan' to exploit us:
It's a hype to be sold to us again and again, because there is big money in it. Without tax payers money and with the obligation to dispose all nuclear waste really responsibly, and being responsible for any damage this technology causes fully at own charge, not a single company will be found doing it.
This includes fusion energy.

“While we have been working decade after decade on developing an incredibly expensive fusion reactor, we are already blessed with one that works perfectly well and is free to all of us: the Sun”
[Hans Joachim Schellnhuber]
 
Maturin said:
Forget it!
You either get more energy density (W/kg), so more miles per kg battery, or less danger. You cannot have both.
Click to expand...
Hmm, isn't it a matter of chemistry and structure? Lithium is highly inflammable, I remember from my school chemistry lessons... but I take your point about energy density, a capacitor will explode if discharged suddenly after all. I think toyota made some claims that their SSB's were less likely to catch fire. There is some discussion of this topic here, the basic claim is that the solid state electrolyte is less inflammable than the liquid/gel electrolyte used in the current generation of batteries.

https://www.topspeed.com/solid-state-batteries-address-fire-explosion-risks/

Whether the SSB's will ever come to market is anyone's guess. All the big companies are working on them though, there have been plenty of announcements in the last couple of years. To my mind its a bit like oled versus lcd for displays.
 
blackbird9 said:
Hmm, isn't it a matter of chemistry and structure? Lithium
Click to expand...
Yes, of course.
I worked a couple of years as a battery's engineer, and know the one thing or the other about Lithium batterys.
It's not easy to put it into a short post, since on the one hand you need to know a bit of electrochemistry to understand some things (you need to read some wikipedia pages yourself on some of the terms I use here), on the other hand it's just pure physics.

Let me try:
Electrical energy (Wh) is the product of voltage and current. So far so good, right? So the energy a battery cell can store comes from two features:
The paring of the galvanic elements (electronegativity) which defines the voltage, and type and density of electrochemical reactive molecules, or as you said 'the structure', which will will decide over a) the amount of charge carriers can be stored (Ah) and b) the max. current the cell can provide and take.

Additionally there is the material that carries the chemicals and conduct the current to the cell's contacts (poles) (some kind of foil), and transport the charge carriers between those (electrolyte).
It's simple to see: the thinner those, the more can be packed into one cell, the more energy it can store, but the more sensible the cell becomes, the more expensive its production is, since you don't want to get a cell's internal short circuit, especially not whil in production (Sony's cell production plant in Japan flared off sometimes between 2005 and 2007 [I forgot] because of that). Cell's internal shot circuit is worst case scenario, because all the charge in the cell is discharged at once, which, depending on the cell, may cause (internal) currents of several hundreds of Ampere, very much more you need for welding, so (very) high temperatures plus according 'side effects'.

But let us look at the other core fundamentals of a galvanic element:
The cell's (rated) voltage comes basically from the pairing of elements of different elecronegativity. To store the most energy electrochemical as possible, you look ideally for the element with the highest positive electronegativity, and the one with the lowest negative one (Those are so almost even, you're almost free to pick any of those) But looking into an accoring table for the positive one will provide you two elements best suitable for a battery cell's cathode:
Lithium and Fluorine (in theory. Forget Flourine directly again! Trying to use this for battery cells is like trying to tame a white shark in pool of blood for your kids having a 'nice ride with a fish'.) Lithium is known to be best material for batteries since the 1920s (or so). The point is, first since the 1980s technology is capable to produce and handle it (as you already said, it's an alkali metal, so very highly reactive - catches fire (metal burn >2000°C, lights almost everything - never try to extinguish a battery fire with water, ever!) even by contact with air.
So with lithium we are already at the max voltage a galvanic element may provide. There is no more to gain.

The second part, the choice and density of molecules which bind the Lithium are the interesting part. Those decide mostly about the energy density (Wh/kg) your cell provides, and the robustness - highest current (charge/discharge) the cell can handle, and how many discharge-recharge-cycles the battery may live until its remaining capicity is reduced so far the battery is not usable anymore.
And here is your main deal you need to make: You cannot have both. You may either have a battery cell capable of high currents (high power cells), or it can store more energy (high storage cells), but then it's less robust, more sensible, can handle lower currents, only.

Anyway if the battery can store more energy, the cell is capabable to cause more damage if it's being destroyed, since the more energy is within it, the more energy can discharge at once. You may reduce this by either a) have the battery store fewer energy, which means the battery stores less energy (Duh! 😜) or b) make the currents on the electrodes flow slower, which means less power in usage, longer times to charge.
You cannot have all - you always need to compromise.

Since the time I started to work on batteries (>25 years ago) there is every year at least on, two, or even more battery cell's chemistries shown as 'THE holy grail - now we have THE ultimate wonder cell...' I don't give a shit on this news anymore for many years.
As I explained above, there is not much margin left anymore anyway, especially not to expect miracle breakthroughs. And anyway a new chemistry doesn't mean anything, especially not there will be a new type of cells. Anything that does it not to market is of no interest at all anyway.

Finally I want to give you a small homework:
Find out some (real, no sci-fi) values of specific energy densities (Wh/kg) for several battery types, such as Lithium based, lead, and maybe NiMH, for capacitors, and for liquid fuels like gasoline, and synthetic fuels, compare those, and summarize you conclusion.

See you (and of course the others, too - I'm always aware my post are not just ment for the addressee, only 😎)
(Damned I wrote that much today here - I wonder if this is appreciated.)
 
Beastie7 said:
From Big Data to Cloud Computing to AI; what do you think is next?
Click to expand...
Next would be some collaps, although it might take long before people realize it.

People mix up intelligence with being able to reproduce large amounts of information. Both have their merits, but IMHO reproducing known knowledge is different from adding knowledge that wasn't here before.

Up till now AI doesn't show other than a very fast fuzzy search engine providing answers in a human readable format. That still makes it not intelligent. It won't solve the problems of today with the knowledge that induced these problems. Maybe AI can lookup data that was hidden from us to answer, but it's still existing data (and knwoledge frustrating egos that hid them for the rest of us).

Because of the deepened confusing between intelligence and mere reproduction, this belief in AI will last for some time. Independent thinkers will know better, but are a minority and will be diffamated for their independent thinking. AI already is citated for corroborating opinions of individuals -- just to make it trustworthy 'because AI also says so'.

There are two things unendless, but we can be sure about human stupidity.

Somehow, some time enough people will realize that independent thinking is essential to answer bigger questions. But it will take time. Long time...
 
I don't think the investors in this technology care whether it's 'intelligent' or not. The motivation from a business standpoint is to develop technologies to automate jobs that are currently done by (large numbers of) humans. Writing code appears to be one of their very first targets, along with 'creative' jobs - music, copywriting, making movies, etc. There are billions in potential savings if you can deskill and/or eliminate the human employees and replace them by machines. Then consider the millions of van, truck and train drivers that can be replaced if self-driving tech can be made to work. There are many other fields that can be considered for automation, for example teachers and doctors. Fully automated AI-controlled dark factories having zero employees, as some companies like Xiaomi already claim to have achieved. Quite who will buy the phones it produces I don't know. Perhaps they will "own nothing and be happy".

It remains to be seen whether any of this will be achieved, but they are throwing billions of dollars at it and they are looking for an ROI. If the ROI doesn't materialise, the bubble will burst.
 
meine said:
People mix up intelligence with
Click to expand...
Why I can only give only one thumb up?
👍👍👍👍
Fully agree. Nothing to add.

blackbird9 said:
The motivation is to a develop technologies to automate jobs that are currently done by humans.
Click to expand...
In the second thought. The first one is always: Can it be sold. Doesn't matter if it's good for anything, nor if it makes the world a better or worse place.
If it can be sold it will be made.

You next part can be seen as the difference of business administration and economists.
The business administrator says:"If I fire all my staff I minimize my cost, so maximize my revenue. I'll be filthy rich!"
The economist says:"If everybody does it that way, you're being broke like everybody else. Because there is nobody left to buy any products anymore."
 
Yep. They are exploring uncharted territory.

Back in the old days they put surplus labour to work building pyramids and temples to maintain the societal stability. Perhaps the modern equivalent will be building a city on Mars.
 
Maturin said:
You may solve afterwards:'yeah, sorry, now we understand. now we solved this problem.' Until the next crash because of:'ohoh, we didn't thought of that.'
Click to expand...
I'm one of those neuro-divergent nay-sayers who easily see the edge cases that fan-boys either don't see or expediently ignore. It makes me somewhat unpopular in software dev circles that I can break most software in ten minutes of use, because my brain is wired to use it "differently" than expected. I've found that even in safety critical development circles the ability to see those edge cases is undervalued by the business interests.

My nightmare is that because the "skynet scenario" was played up in popular movie fiction, the possibiltiy isn't taken seriously, when it should be. I'm not talking about machines rising up against their human opressors. I'm talking about the inherent laziness of humans allowing machines to make critical decisions based on algorithms instead of human intutition and experience.
 
You must log in or register to reply here.
Back
Top