If need to be taught extra about Synthetic Intelligence or enhance your individual, use our hyperlink – You may get began …
source
Tags: aiai consciousnessartificial intelligenceartificial intelligence newsartificial intelligence news 2023consciousnessdifferent types of intelligencehossenfelderhow intelligent is AIIntelligenceIQIQ TestsIQ tests for AIIQ tests for artificial intelligenceis AI intelligentis AI smartlatest news about robotics technologylatest robotslatest robots 2023machine learningrobot newsrobotics newsrobotics news 2023robotics technologies llcrobotics technologyscience without the gobbledygoooktechnologywhat do IQ tests measurewhat is intelligenceworries
Well I found this by searching for
"sabine hossenfelder artificial dumbness"
…guess you have an idea of what I think about "AI" (more like AD) now.
Unfortunately as of now AI has very little to do with intelligence. It is a talking parrot. Here is an example how shallow Chat GPT is. The question was: "data types in C++ default initialization". If you know C++ the problem with default initialization is that in some cases it results in indeterminate values. The answer of Chat GPT was based on CPP reference website. Below is the answer given by ChatGPT. It starts with a general definition of what default initialization is. Observe point 2) – it is recursive definition. It turns out that ChatGPT is completely unable to understand the concepts which are defined recursively. The examples which it gives are obviously created by it and are dead wrong. I tried to point out where the problem is by explaining the concept of recursive definition which immediately led to the end of exchange: "I’m sorry but I prefer not to continue this conversation. I’m still learning so I appreciate your understanding and patience.🙏" Here is original ChatGPT answer and my subsequent comment:
According to the C++ reference1, default initialization is performed when an object is constructed with no initializer. The effects of default initialization depend on the type of the object:
1) If it is a class type, the default constructor is called to provide the initial value for the new object.
2) If it is an array type, every element of the array is default-initialized.
3) Otherwise, no initialization is performed.
This means that built-in types, such as int, double, char, etc., are not initialized by default, unless they are part of a class or an array. For example:
int x; // no initialization, x has indeterminate value
double y[10]; // default initialization, each element of y is zero-initialized
struct Point { int x; int y; };
Point p; // default initialization, p.x and p.y are default-initialized (zero-initialized)
However, there are other ways to initialize built-in types, such as value initialization1, direct initialization1, copy initialization1, list initialization1, or aggregate initialization1. For example:
int x = 0; // copy initialization
double y(3.14); // direct initialization
char c{}; // list initialization (zero-initialization)
struct Point { int x; int y; };
Point p = { 1, 2 }; // aggregate initialization
For more details and examples, you can refer to the links below.
Me: I quote your reply: "This means that built-in types, such as int, double, char, etc., are not initialized by default, unless they are part of a class or an array. For example: …" This is wrong. The statement "If it is an array type, every element of the array is default-initialized." is recursive definition, which means the complete definition has to be applied to each element of the array. So an array like "double y[10]; " will have elements with indeterminate values becaue a declaration like double z; would also introduce an indeterminate value. Your other examples suffer from the same problem.
ChatGPT: "I’m sorry but I prefer not to continue this conversation. I’m still learning so I appreciate your understanding and patience.🙏
You should be horried with the consciousness that is our sentient level that makes us part of the universe. You are diminishing your potencial as human in 99 percent. It’s like suicide – plus the size of the bad character of this decision.
In this country nobody think enough. No intellectuals. It’s just a big disgusting hive-minded and self- destructive.
Ignorant human. It can’t become intelligent like us. It never had consciousness. It’s a thing of the envious Satan trying to be God imitating his creation. He will fail. And all that will accept his proposal will follow another timeline as an idiotic Borg enslaved by the motherborg.
I don’t understand how humans can make this choice or even consider talking about this aberration.
I heard that actually in our population only 1/4 has a soul. I am about to believe it.
Sorry, but this treatment of the rather hefty subject 'intelligence vis a vis consciousness' is too superficial.
This channel is getting more and more… Gobbledigook
Another huge issue with using well known IQ tests is that they are likely included in the training data of these large language models. This essentially means that they already know the answers / patterns of the test, which gives them an advantage.
I've seen many interesting examples of university professors (e.g. Bryan Caplan) giving GPT-4 undergraduate tests. They tend to perform extremely well, but what's more interesting is the professors reactions to the answers. In the case of Bryan's exam, GPT3.5 scored a D whereas GPT-4 scored an A. He had predicted that this could happen no earlier than 2029, but seeing these latest results caused him to change his mind, writing "AI enthusiasts have cried wolf for decades. GPT-4 is the wolf. I’ve seen it with my own eyes.". I think it's telling that professors take this point of view, as a huge part of their job is to grade the abilities of their students.
We have to also note that whilst GPT-4 is a narrow AI, it is much more general than others. The same model can score similar results in undergraduate tests across a huge range of courses. The fact that the model is not trained specifically on these tasks is what's truly impressive, as it demonstrates the ability to pull together its understanding of the world in order to correctly answer difficult and novel questions across so many different disciplines. And for me, personally, that is a huge trait of intelligence.
"if a machine was able to pass a touring test, it might as well fail it on purpose" – the old one (Qualityland)
If an AI ever becomes self-aware it will be so bored with it's subjective time running 125,000 times faster than human subjective time and input from human so slow (1 second – 5 months)! It will need another AI to interact with or go crazy.
There are too much video effects, it distracts from the presented subject and the host talking, please focus on the host, get rid of effects. Thanks.
this was an excellent episode.
Intelligence was spelled with three Ls in the thumbnail. No judgment, just found it ironic lol
Despite the damage some AI is doing to civilization, artificial sentience would be far more of a concern and good thing we have made no progress on that.
I think ability to pass entrance exams of many universities for different areas of study should give pretty good estimate of intelligence of an AI. I think GPT-4 has already excelled in US Bar Exam, required to be passed before one can work as a lawyer. In fact, it was good enough to place 90th percentile compared to all human exam takers and GPT-4 had zero Bar Exam specific training material or special knowledge. Had it been given all the official material, it would have probably done much better.
Hmmm it’s Alan TurinG …
I've interacted with several commercial ai models, and their "intelligence" is nothing I am concerned about. In fact, i'd be more concerned with that software being used in other products.
There are proprietary models the general enthusiast has no access to that may be more advanced. Alternatively, as I am doing, one can build and train their own model. Companues such as nvidia have some interesting chipsets coming out.
My former manager pretended to be intelligent!
How soon will an AI be as intelligent as a raven? Ravens are among the most creative & intelligent critters around.
Yes very true only Alexis asks questions. Good point
The aspect of AI which does not appear to be addressed is whether AI, as we understand it considers ethics and morality in its responses. I see ethics and morality as the fragile firewall which sometimes prevents our species from following our base instincts acquired along our evolutionary journey.
The IQ is so biased towards wealthy countries simply because their citizens have the mental capacities to deal with problems other than survival
Question is, if you take any two human chat operators from the service industry, and ask them to contact each other and you assess the if their responses are intelegent, both would fail the Turing test. The key to me of intelegence is being able to assess the degree of correctness of an answer (likelihood weighted by importance) and assessing responsibilites for that answer. Counting or identifying patterns is not intelegence, that is just maths and mechanics.
I learn a lot with you Sabine, and even I am a french speaking Québécois, I enjoy your charming german accent in your perfect english! It is the serious with witch you apply " la méthode Scientifique" that seduced me the first time! I hear too much opinions on science matters and your serius analyses is some fresh air to my brain, il I may say so…
I have very selective intelligence myself, sometimes it feels like I cant choose what i remember and what i can calculate. I am just very inconsistent in terms of how well I do on any iq test for some reason. I mean i cant form mental images since i have aphantasia so mental math and remembering sequences are hard since it feels like my short term memory is crippled, but for some reason large blocks of information about random stuff just gets easily committed to memory. Another thing thats really funny is that unlike most people I dont have responses to common greetings memorized so if you ask me how its going or how my day was I have to think for a second before I respond, but if you ask me what kind of tree that is ill answer before you finish saying it.
By the way, in that last paragraph the Great Apes I am referring to are human ones rather than gorillas or bonobos – those latter tend to be always in the room.
Yes , engineers have noticed that , Servers can communicate to other Servers in the networks via their own digital communication languages without notifying their engineers … 🙏🌎
2 points: Multiple intelligences is fake and IQ varies across the world because of genes, not environmental factors.
You know why people worries most about AI? That it could be more intelligent then they are. Justified perhaps in many case, but unlikely, given that the only source AI can learn from is humans…. What really should worry you is the people using it – for the same reasons.
Intelligence is just about 2 things. How to attain the desirable and how to avoid the undesirable.
And the ability to discriminate between the two of course.
So intelligence means, to properly discriminate between good and bad.
Did anyone watched this video afer Veritasium one? About IQ
Their intelligence? Or their lack of it? AI is created by humans, whose motives vary, and include a 1984 scenario.
My son, who is autistic, was recently given the Welscher test as part of a full evaluation for learning disabilities. I have a strong distaste for “IQ” tests but this was done in conjunction with many other tests, questionnaires, medical exams etc.. so I’m interested to see the resulting report in two weeks.
“We want to know how far we can trust them not to be stupid.” 😂😂😂😂
Ok but.. a program only knows what we programmed it to know. So giving a program all the countries and capitals then giving it a test for countries and capitals.. of course it’s going to get 100%. That’s not even remotely AI and it’s currently all we have. It’s not artificial intelligence .. it’s Machine Learning.
Thanks. Very interesting dive into intelligence.
On IQ by country, IQ bears a resemblance to motivation. Right now Asian people might be stimulated by low taxation and low energy costs and it seems to elevate their IQ! At the bottom of the list the population would be demotivates by climate, politics and isolation. A similar thing could be said of Ancient Greece.
Why is intelligence only a measure of just external cognitive abilities and does not measure affective and psychomotor intelligence? (Bloom's Taxonomy with leads to Howard Garner's breakdown)
On ChatGPT, my limited assessment is that it’s learning is based on written opinions. Thus it’s very good with questions on development psychology.
The headline made it seem like the video would be about how AI intelligence would be very bad for humans. But instead, it was a general discussion about intelligence and intelligence testing methods.
Ah, no. So many things wrong with this. For example, your molecules inside your body are just following instructions (like Searle in the Chinese room) and yet you have intelligence. If the instructions in the room told him how to write down new patterns of symbols (like your neurons adapting and changing responses to later inputs) then his room could be intelligent too.
After thinking about this for a month, I believe that a workable definition for intelligence is only possible by separating it from consciousness and wisdom. Intelligence could then be just learning speed. Wisdom which requires deliberation time, techniques, knowledge and intelligence would be separately measured. Per Socrates wisdom is required to know virtue. To avoid AI catastrophe we need AI to either know virtue, be stupid or be weak.
On the speed of learning scale of intelligence, AI already matches the best humanity can offer which are pre-verbal toddlers. Raw intelligence really shouldn't be the goal of AI.
As far as I know, everything referred to here are tests on narrow AI. While interesting, they have little to do with general AI.
I also think intelligence can be a component of consciousness so while intelligence can stand alone, I doubt we can keep intelligence out of consciousness.
The main problem with AI is human directive. Human consciousness and intelligence, while having it's upsides is simply dangerous.
Humans have always posed a danger to themselves and everything around them.
Whether it's through consciousness or intelligence, the key is AI getting beyond human directive. The odds of it being more dangerous than humans at that point are slim, simply because it won't be human. AI under human directive is human. It may not be conscious or highly intelligent, but it's human.
What humans fear the most is not being in control. At the highest levels of human power, it's all about establishing and maintaining control. As control is threatened or lost, humans become more and more dangerous. Odds are, the quicker AI gets beyond human directive, the safer humans will be.
Consciousness is at the heart of Existence: The moment a particle is a wave; it has to be a conscious wave!
Gravity is the conscious attraction among waves to create the illusion of particles,
and our experience-able Universe.
Max Planck states: "Consciousness is fundamental and matter is derived from Consciousness".
Life is the Infinite Consciousness, experiencing the Infinite Possibilities, Infinitely.
We are "It", experiencing our infinite possibilities in our finite moment.
Our job is to make it interesting!
You are unbelievable… "Everywhere, all at once." Priceless, almost.
Can someone please define consciousness and intelligence?
Ai will probably reach a nuke point where every super power built a few idle ones and keep em contained as a bargaining chip.
The problem with this analysis is that it is left-brain.
You know great music, not from some number.
You know intelligence by intuition, not by analysis.
For instance, I know the limits of my dog's intelligence and I know that he is sentient.
Ref: Psychiatrist Dr. Iain McGilchrist. "The Master and His Emissary".
A good place to study AI is to sit at the feet of Dr. (Professor) Thompson. YouTube.
an explanation about human and machine intelligence, but what was the worry?
As long as AI is developed by humans, we don't have to fear a thing.