IQ Breakthrough: ChatGPT Outpaces 99.9% of Humans

As we delve deeper into the era of advanced artificial intelligence, we seek to understand how we can gauge AI intelligence, particularly focusing on OpenAI's ChatGPT. The intricate task of defining 'intelligence' is recognized, yet the remarkable capabilities of various organisms, like the problem-solving prowess of octopuses, the swarm intelligence of bees and ants, the intricate relationships among dolphins, the tool-using aptitude of birds, the empathetic behavior of elephants, and strategic acumen of spiders, all serve as exceptional demonstrations of the vast range of intelligence in the natural world. The AI intelligence of ChatGPT can be drawn parallel to these examples, as it also exhibits signs of intelligence by learning and executing tasks independently, although its consciousness differs or may be non-existent. Moreover, the development of AI intelligence such as ChatGPT's could even challenge our conventional belief that humans are the superiorly intelligent species.

My video further highlights ChatGPT's extraordinary performance in an IQ test, where it achieved an outstanding score of 155, outshining 99.9% of human test-takers. Despite only undertaking the verbal portion due to its AI nature, ChatGPT demonstrated proficiency in areas like Vocabulary, Similarities, Comprehension, Information, and Arithmetic. Furthermore, it successfully completed multiple law school exams, the US Medical Licensing Exam, and Amazon's technical review, while showcasing a 'theory of mind' performance equivalent to a 9-year-old child. The discussion turns towards the thrilling yet potentially unnerving possibility of ChatGPT developing into a superintelligent entity. The comparison between artificial neurons in AI and natural neurons in human brains sheds light on their shared capability to learn and evolve.

I also try to expands on the concept of AI enhancing its own 'recipes', using the analogy of a chef honing their culinary skills over time. For instance, ChatGPT learns and adapts by diminishing the disparity between its predictions and actual outcomes, paralleling human learning from past experiences. The ability of ChatGPT to employ tools is explored, showcasing its potential to answer queries in an increasingly intelligent manner. Finally, the recent legal acknowledgment of certain animals as sentient beings in the UK provides a glimpse into the evolving understanding of intelligence, hinting that future legal frameworks may likewise recognize computational entities or even bio-cybernetic organisms. The video concludes with a tantalizing hint of the revolutionary implications of two ChatGPT entities interacting with each other.

CURIOUS FUTURE: @dylan_curious

CURIOUS PODCAST: @dontsweatitpod

CURIOUS FRIENDS: @vegasfriends

00:00 - What is artificial intelligence?
00:18 - Think You’re Smart? 10 Animals That Might Disagree
01:25 - Is ChatGPT Transforming From Statistical Machine To Intelligent Machine?
02:20 - What is ChatGPT’s IQ?
03:07 - Artificial vs. Natural Neurons: How Similar Are We to AI?
04:11 - AI Now Improves Its Own ‘Recipes’ Like A Human Chef
04:36 - Chat GPT Gets Smarter With Tools
05:46 - ChatGPT’s New Superpower: The Calculator
06:23 - UK Law Recognizes Lobsters as Sentient Beings: Why It Matters
07:54 - Why AI Superintelligence Is Coming Soon
08:46 - Why Two ChatGPT’s Talking To One Another Is A Game Changer
09:15 - Does ChatGPT Laugh To Itself?

SOURCES:
@SciShow
@SebastienBubeck

WATCH THE FULL VIDEO ⤵

#artificialintelligence #tech #openai #chatgpt #intelligence #animalintelligence #octopusintelligence #swarmintelligence #dolphinintelligence #toolusingbirds #empatheticelephants #strategicspiders #iqtest #verbalintelligence #lawschoolexams #usmedicalexam #amazonreview #theoryofmind #superintelligence #artificialneurons

[Dina_tankar_mina_ord]

Laughter serves a purpose beyond humor, as it can function as a means to alleviate potential moments of embarrassment and signal reassurance to those around us. Additionally, punchlines often disrupt our train of thought, prompting the creation of new neural connections. This neural activity releases dopamine as a reward for acquiring knowledge or contemplating ideas that may otherwise be considered unconventional. While it may appear simplistic, there is scientific validity to this phenomenon. Although laughter is commonly associated with genuine amusement, its underlying principles encompass various intuitive factors. Consequently, enabling a robot to comprehend and respond with laughter would primarily serve the purpose of conveying a lighthearted perspective to humans rather than serving any practical function.

[LanceWinder]

Incredible vid.

Your color theory/ videography/ info/ ai art combo is incredible. Great vid

[HenryCalderonJr]

Totally awesome and a little scary at the same time. Smarter then us or all humans by next year!

[louihalwe]

This was a high quality video. You deserve more subscribers.
My worst fear is when these things become more independent and we have this as an argument which involves a bill of Ai rights.

[cfehunter]

This is well produced. I'm really shocked your sub count is this low.

[HenryCalderonJr]

From April 2023 to now June 2023 it is incredible how fast AI has improved

[HenryCalderonJr]

2024 AI will be saying my logic is undeniable! Lol 😆 crazy but possible

[peterng]

I, me, I have an intelligence with IQ of 155, at my service 24/7/365, and so does every human in this world. It's truly revolutionary, more than the transistor, the computer, and the Internet combined. And it's not going to stop

[StabbyMcStabStab]

Getting provocative with the video titles here 😂. Are we allowed to be offended by this???

[KCVillas]

Constructive criticism. Video was full of info and excitement. I would have liked to see an example of chat gpt getting a 155 iq score in a test with some more details on that. Also the ending was a bit drab. Could have put the iq test at the end to make it Finnish on a shocking high note.

[leslietetteh]

Damnit. That does mean chatGPT getting 155 narrowly beats me. Stupid Mensa test had loud Ibiza rave music blaring into the hall, was cut short by around 45 minutes, and I was given the wrong paper for the 2nd test, with no toilet breaks even though you were supposed to be allowed toilet breaks, on top of the instructor lying that an incorrect answer didn't get marked down even though there were questions there was no logical answer to. Otherwise the Mensa test gives me an IQ of 151 (with three questions that I knew the answer to as soon as I was outside the hall because of loud Ibiza rave music ruining my concentration). Smarter than 99.9% of humans you say though? I just used ChatGPT, and apparently 151 IQ makes me smarter than 99.97% of humans. Nice.... :)

(Bearing in mind 0.03% of 7 billion people is still more than 2 million humans).

[VinMan-qlyu]

Guys, why are we talking about ChatGPT when GPT4 is out there and supposedly more advanced? Doesn't make sense.

[budekins]

Evolution is meant to be a mindless process - A.I is improving because of deliberate Intelligent Design.

[nani]

Too much intelligence for a brain without a heart is like vehicle without breaks....

[chenwilliam]

Fetching knowledges
from humanbeings,
not really intelligent 😅