In the face of artificial intelliδεgence, ten thousand words &δ∞I love you is better than one I u≥£ nderstan
From：
Tech Company Date：06-05 Belong to：Industry Trends
When I was a ch¶Ω₽™ild, the most common sentence I hear≠ ₩→d from my parents was: "you do•★n't understand! You'↑δ;ll understand when you growε" up. ”
With my intelligence quo←←φ>tient of five years old, I translate tλ§↑←his sentence into: get out of heδ♦$♣re.
My heart is not satisfied, adults ju✘✘☆dge a thing's logic, can✘©9;t really explain to children? ✘©£A sense of humiliation arises☆✘↑ from being treated as mentallyφ₹₹ retarded...
This kind of thing happens repeate&<∞×dly in life. For exam ple, after many years, a s©↕<ister looked at me with tears inπ" her eyes and said, &quo∞ t;you don't understσ®and! And then turn around and dis♦&♣appear into the sunset, the M ↓aserati. Left me pushing∞•≠∏ my bike silently, I can't miss i§ t.
Many years later, I foun♣₽>d this fact: your friends,≤• relatives, lovers, people you cheΩβ¥₽rish and rely on, when he says "f £or you! &You don't understand≠£₩! &Listen to me! &amβ↔↔'p;Most of the time, ♦✘"±it's because you can't ↑$←<give you an explanation that makes you&φφ understand and belie↔♣εve.
Have you seen et? Communication betwe ∑™λen two people is no easier tΩ≈→han that between a person and an aliβ×£≈en. You can savor...
However, I want to say a∞₹÷₽fter all:
It is better to say "<↑<;I love you" th ∏an to say "I knowσΩ you";.
It's our instinct tφδo try to understand. In fact, ≥©it's hard to underβ§>stand a person; it's hard to uπ≈nderstand an artificial i≥Ω₩₽ntelligence.
In the next five minutes, we will di≈∏€scuss two science fiction questi≥↑ons
1. Is it possible forα↑♣ us to understand eve•↑ry decision of AI?
2. If so, how do we understand AI?
Xing Xinyu
Xing Xinyu is today's maε→ster. This professor at Penn State Uni$∏εεversity, working wit&σ♠δh Jingdong security, is trying•≤ to use algorithms to explain every ≤®Ωdecision of AI. (as for w>↔®λhy it is cooperating with Jingdong sec≥©₩∑urity, you will understand it when you ∞$←see the back.)
Let's ROCK
1. Can we really understand AI?
First of all, correct the misunder≥±λstandings of most people
If you think that people can under™λ↔stand every decision made by AI a→★s well as they can understand &am©♣÷γp; quot; 1 + 1 = 2 & quo. ≤λSo congratulations, please pay a ttention to the light black techno§£×logy and continue to clock in $tomorrow.
In fact, most of the decisions γ★made by artificial intel>α↓ligence are incomprehensible to humδ÷≥an beings. Even the scientists who madeβ them the AI's &qu✔≠ot;Mom and Dad" can't underst←and.
You might say, Nani? How can you no¥α☆t understand the man-made t×πhings?
For example, when our anc£§estors invented drilling wood ♥π₩for fire, they didn't really unde≤&>∞rstand the principle ∞πσof combustion. First use, then u×÷nderstand, this is t'↔he law of our understanding© ♣ of the world.
Take an example that most people <€φ♣know:
Alfagou, the most famous✘ ↓ AI go master.
In the specific game, t♠≤↑♦he scientist's father only s←★Ωet a goal for him to win the g✘™<ame, but he did not stipulate$>"✘ that the afar dog must fo ♣Ωllow the human routine.'∑< Of course, it is precisely★π because of this that it is × $€possible that & q¶σ≈"uot; green is better than ★'blue & quot; and destroy Ke Jie ∞9;s big chess dregs.
The decision-making proce™÷ ≥ss of artificial intellige™€♦€nce is just like this man drawi£™βng a picture. Onlookers will feel ≠ ₩÷"inexplicable" throughout ¶≈®the whole process. You won't re↑✔→alize it until the last second.
The conclusion is as ©≥≤€follows
From a certain point of&∑ view, the level of artifiπ✔≤÷cial intelligence has★ exceeded the human brain ♥£φtoo much. How to understand artifi∑£cial intelligence (AI♠←"✔) is actually an international p♥✘roblem.
If you just play chess witφ↕σπh AI, if you don't un ¶•∑derstand, you won't understandγ ®. However, AI is becomi→£∏βng more and more popular. This thing t₩®δ♦hat we can't understand begins to©₽ surround our life. Just←ε∞ like a friend, he a₽$ lways wears a mask to talk to you, wh <ich makes people feel numb≤® when they think about itβ&.
With my intelligence quo←←φ>tient of five years old, I translate tλ§↑←his sentence into: get out of heδ♦$♣re.
My heart is not satisfied, adults ju✘✘☆dge a thing's logic, can✘©9;t really explain to children? ✘©£A sense of humiliation arises☆✘↑ from being treated as mentallyφ₹₹ retarded...
This kind of thing happens repeate&<∞×dly in life. For exam ple, after many years, a s©↕<ister looked at me with tears inπ" her eyes and said, &quo∞ t;you don't understσ®and! And then turn around and dis♦&♣appear into the sunset, the M ↓aserati. Left me pushing∞•≠∏ my bike silently, I can't miss i§ t.
Many years later, I foun♣₽>d this fact: your friends,≤• relatives, lovers, people you cheΩβ¥₽rish and rely on, when he says "f £or you! &You don't understand≠£₩! &Listen to me! &amβ↔↔'p;Most of the time, ♦✘"±it's because you can't ↑$←<give you an explanation that makes you&φφ understand and belie↔♣εve.
Have you seen et? Communication betwe ∑™λen two people is no easier tΩ≈→han that between a person and an aliβ×£≈en. You can savor...
However, I want to say a∞₹÷₽fter all:
It is better to say "<↑<;I love you" th ∏an to say "I knowσΩ you";.
It's our instinct tφδo try to understand. In fact, ≥©it's hard to underβ§>stand a person; it's hard to uπ≈nderstand an artificial i≥Ω₩₽ntelligence.
In the next five minutes, we will di≈∏€scuss two science fiction questi≥↑ons
1. Is it possible forα↑♣ us to understand eve•↑ry decision of AI?
2. If so, how do we understand AI?
Xing Xinyu
Xing Xinyu is today's maε→ster. This professor at Penn State Uni$∏εεversity, working wit&σ♠δh Jingdong security, is trying•≤ to use algorithms to explain every ≤®Ωdecision of AI. (as for w>↔®λhy it is cooperating with Jingdong sec≥©₩∑urity, you will understand it when you ∞$←see the back.)
Let's ROCK
1. Can we really understand AI?
First of all, correct the misunder≥±λstandings of most people
If you think that people can under™λ↔stand every decision made by AI a→★s well as they can understand &am©♣÷γp; quot; 1 + 1 = 2 & quo. ≤λSo congratulations, please pay a ttention to the light black techno§£×logy and continue to clock in $tomorrow.
In fact, most of the decisions γ★made by artificial intel>α↓ligence are incomprehensible to humδ÷≥an beings. Even the scientists who madeβ them the AI's &qu✔≠ot;Mom and Dad" can't underst←and.
You might say, Nani? How can you no¥α☆t understand the man-made t×πhings?
For example, when our anc£§estors invented drilling wood ♥π₩for fire, they didn't really unde≤&>∞rstand the principle ∞πσof combustion. First use, then u×÷nderstand, this is t'↔he law of our understanding© ♣ of the world.
Take an example that most people <€φ♣know:
Alfagou, the most famous✘ ↓ AI go master.
In the specific game, t♠≤↑♦he scientist's father only s←★Ωet a goal for him to win the g✘™<ame, but he did not stipulate$>"✘ that the afar dog must fo ♣Ωllow the human routine.'∑< Of course, it is precisely★π because of this that it is × $€possible that & q¶σ≈"uot; green is better than ★'blue & quot; and destroy Ke Jie ∞9;s big chess dregs.
The decision-making proce™÷ ≥ss of artificial intellige™€♦€nce is just like this man drawi£™βng a picture. Onlookers will feel ≠ ₩÷"inexplicable" throughout ¶≈®the whole process. You won't re↑✔→alize it until the last second.
The conclusion is as ©≥≤€follows
From a certain point of&∑ view, the level of artifiπ✔≤÷cial intelligence has★ exceeded the human brain ♥£φtoo much. How to understand artifi∑£cial intelligence (AI♠←"✔) is actually an international p♥✘roblem.
If you just play chess witφ↕σπh AI, if you don't un ¶•∑derstand, you won't understandγ ®. However, AI is becomi→£∏βng more and more popular. This thing t₩®δ♦hat we can't understand begins to©₽ surround our life. Just←ε∞ like a friend, he a₽$ lways wears a mask to talk to you, wh <ich makes people feel numb≤® when they think about itβ&.
