The potential effect of Artificial Intelligence on civilisation - a serious discussion

We've looked at the possibility of AI going wrong, but what if some lunatic designer deliberately programs it to be bad, or it gets hijacked/virus'd etc.
 
Harking back to those archaeologists and eg cuneiform...

Re-reading 'Ark before Noah', about the even-earlier Gilgamesh version recently revealed.
Spoiler: It was round...

Author, IIRC, mentioned that there are scant few, generally elderly scholars able to translate cuneiform, and very many inscriptions yet to be translated / reported.

IMHO, being able to point phone at a random inscription as if at QR code and get a 'first pass' translation, preferably with 'degrees of confidence' mapped, would be a mercy...
--
Sorta tangential, I remember when tables of logarithms were deprecated, followed by slide-rules, then basic calculators...

Per OBB's comment about calculus etc, I hit a 'glass ceiling' at 'non-trivial', had to abandon aspirations to PhysChem, ChemEng, Oceanography...

For the record, the most actual use of chemistry from my formal education was to explain to a hapless lab-tech how soap worked and why it was essential in our auto-analysers' mobile phase. Happens that had been the first question on Chem GCSE exam when I was 16...

Everything else, I gleaned from side-streams such as hobbies, my eclectic curiosity or learned 'on the job'...

And 'structure modelling' ? Who in their right mind would analyse anything more complex than very simplest 'pin-joint' structure by manual calculation using matrices ? I helped out a young colleague doing OU course who'd come to grief on such. He'd attacked his 'Statics' assignment thrice over a week, got different results each time. Other students fared worse. I went back to 'first principles', broke forces down to orthogonals using Sin/Cos, had it done in ten minutes. Literally, 'On the back of an envelope'...

But, and a very, very big 'but', the matrix approach --Mastered-- permitted prompt computation by PC. A dozen factors ? A thousand factors ? 25 k factors ? 250k factors ? A 'Bazillion' ?? Whatever !! Like a CGI render, complexity just takes a little longer: Run over coffee break, lunch, over-night. On this CAD-Tower PC, or my network-render 'Box' nearby. Or, if serious money may be traded for time, an off-site 'render farm'...

AI is like any other 'power tool', such as my formidable plunge-saw: Requires 'Due Care', has a steep learning curve, and will exact a Darwinian toll on the foolish and/or unwary...
 
Apophenia said:
Justo Miranda said:
GTX said:
A useful development IMHO:

www.heritagedaily.com

Archaeologists use artificial intelligence (AI) to translate 5,000-year-old cuneiform tablets

A team of archaeologists and computer scientists have created an AI program that can translate ancient cuneiform tablets instantly using neural machine learning translations. - HeritageDaily - Archaeology News
www.heritagedaily.com www.heritagedaily.com

Would be interesting to see what is learnt as a result.
Click to expand...


Not too much, for the most part these tablets contain accounting records ...
Click to expand...


Those "accounting records" are pure gold for Assyriologists and other Ancient Near East studies. Most ANE trade routes have been uncovered due to the translating of such 'mundane' record-keeping. (And, yes, that is important for understanding the origins of human civilization.)

Remember, for example, that archeology is a study where invaluable details are found in palimpsest manuscripts - what we'd now call notes scribbled on recycled scrap paper. And archeologists spend much of their time excavating middens. You may not want to dig in refuse heaps but I, for one, am grateful that archeologists do.

Sure, the cataloguing of receipts for the ancient textile and tin trade to Anatolia is not for everyone. Then again, none of my neighbours would agree that obsessing over old airplanes is a good use of one's time either ...
Click to expand...
https://www.deseret.com/2014/1/24/2...british-museum-contains-building-instructions

https://www.theguardian.com/culture...an-tablet-noah-ark-constructed-british-museum

https://www.theatlantic.com/technology/archive/2014/01/noahs-ark-round/283335/

https://historyofinformation.com/detail.php?id=4060
 
edwest4 said:
Experts, amateurs. What's the difference? Take this site.

Seriously, I 100% decry 'armchair experts,' who know very little, posting anywhere. Would you go to an amateur doctor?
Click to expand...
That’s a whole discussion all by itself. Nothing is more individual than medicine…or generates more retractions.

My Dad, a man who never finished grade school…couldn’t convey his level of pain to doctors who made their minds up he just had arthritis.

The MRI results my family got literally the day we put him in the ground showed not just a “knot” on his aorta…but a massive tumor on his spine.

I kick myself for doing what the doctors said in trying to get him to walk every day.

Lots of folks have to go to multiple doctors to be heard.

Outsiders have a role to play (sometimes) and they aren’t all cranks who think they know better than Einstein.

Sometimes, Ph.D might as well stand for “post-hole digger.”

I seem to remember a news item where a faker pretended to be a physician…where some patients bragged on how sympathetic and caring he was…versus their usual doctors with bad attitudes.

In medicine, quackery begins where sympathy ends.
 
Last edited:
Arjen said:
I may be guilty of slandering an entire profession here, but I recall reading somewhere - unfortunately I can't remember where - that of all university students, law students were the hardest to convince of the statisticians' position in the three doors problem.
Click to expand...
Perhaps they were looking for a fourth door out?
 

Attachments

  • 1476455249_shark-8.jpg
    1476455249_shark-8.jpg
    20 KB · Views: 7
  • padre-es-detenido-por-encerrar-a-su-hijo-en-una-jaula-durante-mas-de-veinte-anos-655x368.jpg
    padre-es-detenido-por-encerrar-a-su-hijo-en-una-jaula-durante-mas-de-veinte-anos-655x368.jpg
    37.5 KB · Views: 7
Forest Green said:
GTX said:
www.abc.net.au

ChatGPT-style app to be trialled in schools to teach AI safety

Education Minister Blair Boyer says it is essential that students learn to use AI safely before they graduate. The trial is set to run for two months, after which a decision about a broader rollout will be made.
www.abc.net.au www.abc.net.au
Click to expand...
Using AI to teach AI safety? Hmmm... there's an obvious problem there.
Click to expand...

It's a marketing move. Illegal drug dealers sometimes hand out free samples in the hope of getting new drug users. This is similar. Get young people exposed to AI, which allows them to "understand" it, and hopefully, they become regular users.
 
edwest4 said:
And now a message from Microsoft:

"We are financing the AI Apocalypse."

What. A. Joke.
Click to expand...
An apocalypse from time to time can be something necessary for human evolution, at present the fittest keeps with their taxes the less fit that are increasingly numerous because nature can no longer eliminate them as easily as before.
 
Another message from Microsoft:

Hi Everyone,

Are you scared? We sure hope you are. We need you to be scared so you can use devices that run on our software a lot more. It's the weekend in the U.S. We need you to be scared about Extreme Heat. We hope you don't notice that this record breaking heat is unofficial. It's not like we don't have global temperature monitoring and satellites. Next year in the U.S., is an election year. We have a lot of friends in the U.S. government and they need you to be scared about something. That way, they can offer you solutions to what you're scared about. We need you to vote for our candidates because they'll come up with ideas to get rid of the (unofficial) extreme heat with equipment that runs on our software. We have no plans of dying. We want to live to a ripe old age and spend our billions of dollars.

We now return you to our regularly scheduled mass shootings.

Microsoft - Official Member of the Ruling Elite.
 
publiusr said:
That’s a whole discussion all by itself. Nothing is more individual than medicine…or generates more retractions.

My Dad, a man who never finished grade school…couldn’t convey his level of pain to doctors who made their minds up he just had arthritis.

The MRI results my family got literally the day we put him in the ground showed not just a “knot” on his aorta…but a massive tumor on his spine.

I kick myself for doing what the doctors said in trying to get him to walk every day.

Lots of folks have to go to multiple doctors to be heard.

Outsiders have a role to play (sometimes) and they aren’t all cranks who think they know better than Einstein.

Sometimes, Ph.D might as well stand for “post-hole digger.”

I seem to remember a news item where a faker pretended to be a physician…where some patients bragged on how sympathetic and caring he was…versus their usual doctors with bad attitudes.

In medicine, quackery begins where sympathy ends.
Click to expand...
I feel your pain, I'm sorry for your loss.

Worked with doctors in the acute side of medical care and can confirm that a 'lot' of doctors see patients as subordinate members of the species with arrogance as their major.

Numerous failings that cannot be unseen and even as an advanced paramedic, when I try to convey the issues I have, they fall back on classic diagnosies that have no relation to what I describe. Next 'Quack' to laugh at my pain will be going astraight to somewhere comfortable and safe

I am not saying all doctors are bed but frankly they do not understand 'informed consent' which is the legal standard any more than they understand the phrase, "Fix it don't play with it".
 
One of my fears is that the AI you want may decide that it doesn't want you.
 
Foo Fighter said:
Worked with doctors in the acute side of medical care and can confirm that a 'lot' of doctors see patients as subordinate members of the species with arrogance as their major.

Numerous failings that cannot be unseen and even as an advanced paramedic, when I try to convey the issues I have, they fall back on classic diagnosies that have no relation to what I describe.
Click to expand...
publiusr said:
Lots of folks have to go to multiple doctors to be heard.
Click to expand...
Oh man have I lived that.
And so have quite a number of people in the disability and the incurable illness communities.

Not to send things off on a tangent, but 2 things come to mind from y'all's comments,

One,

As you were first learning about the differential diagnosis process, you were likely exposed to the maxim, “When you hear hoofbeats, think horses, not zebras.” In other words, when diagnosing patients’ signs and symptoms, remember that many diagnoses (horses) are common, and the chance that a patient has a rare diagnosis (a zebra) is small. This wise phrase is attributed to Theodore Woodward, MD (1914-2005), a respected physician, researcher, and teacher at the University of Maryland who also was a charter member of the Infectious Diseases Society of America.
Click to expand...

Two,

Think ZEBRA!
April is National Primary Immune Deficiency Awareness Month!
In medical school many doctors are taught the old saying “when you hear hoof beats, think horses, not zebras”, which means that doctors should consider the most likely possibility first when thinking of a diagnosis. However, primary immune deficiencies are rare (like zebras!), thus the Immune Deficiency Foundation started PID Awareness Month to remind patients, families and doctors that not all diseases are horses… some are zebras!
Click to expand...

See also:

med school versus life.jpg
 
Last edited:
Something common with failure in any industry is the domino effect.

If an aircraft comes down it us normally a series of failures to act on a specific problem such as switcvhing off the wrong engine, as has happened.

Doctors are just as likely to make a diagnosis based on instant diagnosis via It must be X because that is the mnost common problem. Not all infection will lead to a raised internal temperature for example but gp's/family doctors act as though they do simply because it is the most common effect.

Crohn's disease will have consequences like a difficulty breaking down vitamin B12, add to this the prescription and antacids will lead to nil extraction of B12 which affects the immune response.

This will lead to infections of mucosal tissue, leading to an increase in gum and sinus infection, this will cause post nasal drip and often times a sinus between the sinus and the gums. It also means you are swallowing infected mucus. WHich leads to aeropahgia and problems with hiatus herniation.

So you have a medical condition which leads to all sorts of problems including dental infection which dentists will not treat as it is 'Medical' in origin and doctors will laugh at saying "Just get the f'ing teeth ripped out"!

So, medical and technical failures come down to arrogance and ignorance on the part of those involved, just try getting anything done about it unless the problem kills a lot of people at once.
 
If you make friends with a doctor, he will tell you which are the good ones, the ones who take care of his mother. This trick has worked well for me.
 
Australian Air Power Today May 2020-Aritificial Intelligence 04-13
 

Attachments

  • page_1.jpg
    page_1.jpg
    242.3 KB · Views: 3
One exercise I'd like to give to two sets of AI...one air-gapped like Watson on JEOPARDY and the other plugged to the internet...would be how best to send payloads to space.

The air-gapped stand-alone would not have any aerospace info at all...but would have to start from scratch.
 
In the news
phys.org

Digital technology in government: Over-promised and under-delivered, says study

Politicians and tech bosses have spun a "Digital Enchantment" fairytale, which does not present a realistic or helpful view of digital technology, according to new research from the University of Birmingham.
phys.org phys.org
techxplore.com

Do androids laugh at electric sheep? Study challenges AI models to recognize humor

Large neural networks, a form of artificial intelligence, can generate thousands of jokes along the lines of "Why did the chicken cross the road?" But do they understand why they're funny?
techxplore.com techxplore.com
techxplore.com

Computer system based on light could jumpstart power of ChatGPT-type machine-learning programs

ChatGPT has made headlines around the world with its ability to write well-done essays, emails, and computer code based on a few questions from a user.
techxplore.com techxplore.com
phys.org

Research team takes neuromorphic computing a step forward

Neuromorphic computers do not calculate using zeros and ones. They instead use physical phenomena to detect patterns in large data streams at blazing fast speed and in an extremely energy-efficient manner.
phys.org phys.org
techxplore.com

AI can ask another AI for a second opinion on medical scans, research shows

Published recently in Nature Machine Intelligence, research addressed the limited availability of human annotated, or labeled, medical images by using an adversarial, or competitive, learning approach against unlabeled data.
techxplore.com techxplore.com
phys.org

New AI model outperforms traditional methods in predicting central Pacific El Niño

In a recently published study, researchers from the Institute of Atmospheric Physics of the Chinese Academy of Sciences have developed an artificial intelligence (AI) model using deep learning algorithms that outperforms traditional dynamic models in predicting the development and pattern of El...
phys.org phys.org
techxplore.com

Researchers develop machine-learning technique that can efficiently learn to control a robot

Researchers from MIT and Stanford University have devised a new machine-learning approach that could be used to control a robot, such as a drone or autonomous vehicle, more effectively and efficiently in dynamic environments where conditions can change rapidly.
techxplore.com techxplore.com

For the polymaths:
phys.org

Study proposes combining continuum mechanics with Einstein field equations

Albert Einstein's general theory of relativity is a landmark in our understanding of the universe. It gave rise to the notion of a spacetime continuum against which all physical phenomena play out. But over the decades, it has inspired many questions that have yet to be answered: How can...
phys.org phys.org

Monte Carlo sim breakthrough
phys.org

Physicists achieve breakthrough in Monte Carlo computer simulations

Researchers at Leipzig University have developed a highly efficient method to investigate systems with long-range interactions that were previously puzzling to experts. These systems can be gases or even solid materials such as magnets whose atoms interact not only with their neighbors but also...
phys.org phys.org
 
The UK intelligence agencies are lobbying the government to weaken surveillance laws they argue place a “burdensome” limit on their ability to train artificial intelligence models with large amounts of personal data.

The proposals would make it easier for GCHQ, MI6 and MI5 to use certain types of data, by relaxing safeguards designed to protect people’s privacy and prevent the misuse of sensitive information.

Privacy experts and civil liberties groups have expressed alarm at the move, which would unwind some of the legal protection introduced in 2016 after disclosures by Edward Snowden about intrusive state surveillance.

The UK’s spy agencies are increasingly using AI-based systems to help analyse the vast and growing quantities of data they hold. Privacy campaigners argue rapidly advancing AI capabilities require stronger rather than weaker regulation.
Click to expand...
www.theguardian.com

UK spy agencies want to relax ‘burdensome’ laws on AI data use

GCHQ, MI6 and MI5 propose weakening safeguards that limit training of AI models with bulk personal datasets
www.theguardian.com www.theguardian.com
 

Similar threads

Back
Top Bottom