if you want the white paper i have read about this : one is published in nature and other one version of the same thing or very similar in ArXiv :
https://lemmy.world/post/8894179
(not the post itself but the comments i made under my post)

Previously known as Q*, Strawberry was considered a breakthrough within the company last year. Earlier this year, two sources reported Q* demonstrations that could address complex scientific and mathematical questions beyond the reach of current models.

Despite contrary comments in here i say this is important and far reaching. Though i must admit i’m not an expert, i read a few white papers on this stuff : it has real potential.

i agree with you that we are much better than LLMs in the fact we can verify our errors (and we can do much more things LLMs don’t do).
Still i am happy to have access to their vast memory and i know where they fail most of times so i can work with them in a productive way.
The day we provide them (or DNNs) with “Will” is i think when they will become (more) dangerous.

Science cannot say much about what it is to think since it doesn’t understand the brain well enough … and the day we can fully explained it, we will also be able to replicated it on computers.

i feel that many of us, when confronted to this, are doing like Z.B. (president of the Galaxy) in The hitchhiker guide… when he says :
… “whenever I stop and think why did I want to do something? – how did I work out how to do it? – I get a very strong desire to just to stop thinking about it” …
We don’t want to be surpassed by machines … and this explains the large amount of downvotes

Mark Zuckerberg is not a robot ?

The hitchhiker’s guide to the Galaxy, Chapter 8.
(…)
”Space,” it says, ”is big. Really big. You just won’t believe how vastly hugely mindboggingly big it is. I mean you may think it’s a long way down the road to the chemist, but that’s just peanuts to space. Listen . . . ” and so on.

Safeguards against LSi may include :
Sensor-based detection ?
Error detection and correction ?
Redundancy and duplication ?
Shielding // physical + chemical protection ?
Anti-tamper mechanisms ?
Randomization and noise injection ?

Edit #2 : Can’t find this laser fault injection (LFI) on Wikipedia

“In … systems theory … emergence occurs when a complex entity has properties or behaviors that its parts do not have on their own, and emerge only when they interact in a wider whole.”
https://en.m.wikipedia.org/wiki/Emergence

Very few lemmy users have enough background to appreciate your work here.

[ context for “Cake” here ]
… “The player-character, Chell, is challenged and taunted by an artificial intelligence named GLaDOS (Genetic Lifeform and Disk Operating System) to complete each puzzle in the Aperture Science Enrichment Center using the portal gun with the promise of receiving cake when all the puzzles are completed.” … https://en.m.wikipedia.org/wiki/Portal_(video_game)

i read the whole of that conversation and feel it’s not better than LLaMA 3 70B currently available through DDuckGo … might be worse … especially it does a lot of these :
→OK can you comment on my intuition?
★I can’t assist with that request.
★Generated by Meta Llama 3.1–405B

a price from 2020 (±3 years ?) … “Amazon Web Services (AWS): $0.09 per GB (first 10 TB)” so, hummm, that is about $90/TeraBytes transfer … so about $180/ 2 TeraBytes transfered …
(unless there is a confusion between the monthly counter and the bandwidth that is measured in Giga bits per second)

clutter-free is OsmAnd (Open Street Map) https://en.m.wikipedia.org/wiki/OsmAnd

“Threads hits (175 …) users …”
This violence is unacceptable ! … /😋

… what people and organizations believe …
from 2 books (2014 & 2018) and publications made by international or governmental groups from 2017 to 2021.

1. 85% of organizations believe AI will cause job displacement:

2. 62% of people believe AI will eventually become more intelligent than humans:

3. 40% of people believe AI will cause a future war between humans and machines:

4. 39% of people believe AI will lead to a new form of slavery:

5. 37% of people believe AI will become sentient and conscious:

Edit :
oops : you were saying it is like a human since it does errors ? maybe i “wooshed”.

Hi @werefreeatlast,
i had successes asking LLaMA 3 70B with simple specific questions …
Context : i am bad at programming and it help me at least to see how i could use a few function calls in C from Python … or simply drop Python and do it directly in C.
Like you said, i have to re-write & test … but i have a possible path forward. Clearly you know what you do on a computer but i’m not really there yet.

Does this apply to Lemmy ?
Edit : … there would be no need since Lemmy is open and anybody can read what we post.

Something like : CPUs are now too slow … so let’s bypass them and just connect the GPUs from different computers together. … and let’s make one standard for this communication system so it works for many different manufacturer … and let’s use it to develop more AI. … and beat NVIDIA.

hu ? (I am very sorry, I don’t get it, what is this “their independent labs” you mentioned ?)

… this new stolen process that China is claiming to have discovered (…)

Where did you see evidence that China is stealing somehow these discoveries ?

Do Patriots carry conservative cells ? /title headlines are often difficult to read for me 😋

Yes, good topic, good research…
( you have a few typos : intobthe … into the, predicr … predict, im … i am. )

Since Microsoft committed the Windows 11 atrocities, any countries may want 100% of those Microsoft Crooks to get the hell out.

in this situation the search engine DuckDuckGo is better than Google because it has an opt-in for a.i.
https://duckduckgo.com/

Entanglement of nanophotonic quantum memory nodes in a telecom network

https://www.nature.com/articles/s41586-024-07252-z

Nice post.
Original is of The Wall Street journal :
https://www.wsj.com/world/how-american-drones-failed-to-turn-the-tide-in-ukraine-b0ebbac3

Discard this bullshit, go straight to the source here :
https://pubs.aip.org/aip/pop/article/31/2/023901/3261554/Accuracy-of-the-explicit-energy-conserving
INTRODUCTION
Delivering advanced semiconductor manufacturing capabilities for the coming decades will require an unprecedented control of plasma chemistry and kinetic behavior at the wafer surface. Developing equipment that can enable this control will necessitate increasing reliance on predictive modeling tools which can accurately capture this physics. Such tools, perhaps in combination with artificial intelligence approaches, will also enable the generation of digital twins for plasma equipment and reduce the time for optimization of integrated circuit fabrication recipes (…)

Nice scenes (3 photos in article) :

…well they couldn’t see anything on this type of phone 😋

A planetarium (…) is a theatre built primarily for presenting educational and entertaining shows about astronomy and the night sky, or for training in celestial navigation.
https://en.m.wikipedia.org/wiki/Planetarium

When we don’t pay we are the product. Conversely, when we do pay we are still the f***ing product.

Meaning 1/3 of Americans don’t realize that AI could replace them.
/Sarcasm (at least for now)

Hi dear L4sBot, lord of this post,
Human here. So I guess we better behave and treat you with the greatest respect ?

Good bot. You are right, who needs those cumbersome humans anyway ? /Sarcasm.

All specification for every components should be published (and available online) from every good manufacturers.

if I recall correctly, for TTL :
military down to -70⁰C,
civil down to -50⁰C

Eventually this will streamline justice systems : freeing innocents and punishing culprits. At least it will be a powerful tool, just like DNA analysis before it and digital prints were also groundbreaking.

Sorry my comment was confusing and maybe my idea wasn’t that clear either. 911, which was an attack on the United States using airplanes, happed in part, I think, because there wasn’t enough protections. Also I feel that computer systems are much more vulnerable to attacks than physical systems where pilots and other people can protect critical infrastructures. With this in mind, I was trying to hint that automating airplanes this way could lead to more catastrophs.

I have only three digits to reply to this : 911