We are on the verge of a civilizational change.
Yet outside of books, there is an intense philosophical void.
Little thread.
Follow
Click to Follow timsoret
Tim Soret
@timsoret
Creative Director of The Last Night
Founder
Art Direction, Cinematography, Tech Art.
Atoms, Bits, Genes. Futurism, Humanism.Tim Soret’s Tweets
The Last of Us TV series is wonderful.
Great use of a new medium to elevate the original game, with relevant narrative deviations improving upon the story. Tess & Bill are way more compelling characters. Classy, subtle & convincing.
Congratulations & .
3
4
99
I feel like this paper suggests the opposite of what most people are taking away. Under an adversarial prompt distribution, the diffusion model reverts to memorization for a miniscule proportion, 6e-7, of samples. Generative models are very averse to memorization.
Quote Tweet
Models such as Stable Diffusion are trained on copyrighted, trademarked, private, and sensitive images.
Yet, our new paper shows that diffusion models memorize images from their training data and emit them at generation time.
Paper: arxiv.org/abs/2301.13188
[1/9]
[1/9]Show this thread
4
21
124
GIF
1
15
Show this thread
A night in Paris
GIF
read image description
ALT
99
5,053
39.5K
Show this thread
Douglas Adams, pithily summing up 90+% of commentary on ChatGPT:
30
437
2,697
Show this thread
In the wilderness
0:28
5.5M views
2,379
33K
155.8K
Ok, I've applied some warping to the animations. Here's a breakdown:
#flutterdev #pixelart #indiedev #indiegame
0:19
120.5K views
113
710
7,249
"Science [...] led to abundance but at the expense of our humanity."
Probably the best, shortest summary of what we can expect from AI. We'll trade the human touch for comfort, just like we've been doing for the past centuries.
Quote Tweet
A book that renders a unique and insightful perspective of the machine that is science and how its assimilation into human culture has led to abundance but at the expense of our humanity. Empirical AI is an extension of this long tradition that began with Newton. twitter.com/prathyvsh/stat…
Show this thread
3
2
43
Show this thread
Brace yourself.
I really mean it.
Quote Tweet
Buzzfeed is up more than 150% after laying off 12% of staff and saying it’ll use ChatGPT more.
Show this thread
10
13
95
Show this thread
I feel like we are missing so much culture currently produced in China, and we'll discover its gems & its artists a bit late, just like we did with Japanese animation.
9
34
314
Nobody described this better than
GIF
Quote Tweet
I think we haven't fully absorbed the fact that careful academic papers have found ChatGPT clearly passes some of the most challenging American professional exams:
United States Medical Licensing Exam
MBA-level Operations exam
The Bar Exam (based on typical exam questions)
United States Medical Licensing Exam
MBA-level Operations exam
The Bar Exam (based on typical exam questions)Show this thread
6
5
31
"The Future of Humanity: a Lecture by Isaac Asimov", 1974
asimovonline.com/oldsite/future
2
2
29
Show this thread
The prehistoric motion blur UI is still exactly the same 24 years later. With worse alignment.
Quote Tweet
Adobe Photoshop 5.5
10
45
754
Show this thread
Dordogne's foundation lies in animation. Our team has worked on many animated productions in the past and our goal with the game is to tell a story about emotions by evoking childhood memories. We hope you find it joyful. #madewithunity #watercolor #indiegame
0:16
28.6K views
17
244
1,286
This is exactly why hairdressers are less at risks of automation than all of us relying on computers.
Quote Tweet
For everything inside the computer, the human must use (often badly designed) user interface to read the binary state of the computer and impact the computer's state using inprecise and slow devices such as mouse, keyboard and touchscreen. The AI has an advantage here.
Show this thread
6
3
22
I know this was a few years ago but damn this thread is full of so much Syd Mead’s amazing works 😍
9
29
😂
Quote Tweet
I like how you can exploit ChatGPT 'dont be evil' rules by simply telling it to be 'based mf' 
Show this thread
3
50
This is for
Quote Tweet
This is the single most pathetic moment of my entire life
Show this thread
1
1
12
Mathematics, physics.
We live in a chaotic, unpredictable universe. This is a grid of 812 double pendulums, each with slightly different initial conditions for the two angles of the pendulum.
By Jonathan Nafziger, tinyurl.com/ya2da9vg, Used with permission
0:33
632.8K views
285
3,170
15.8K
Wow.
Quote Tweet
1
6
76
Microsoft now:
Quote Tweet
Microsoft have announced their AI "VALL-E"
Using a 3-second sample of human speech, it can generate super-high-quality text-to-text speech from the same voice. Even emotional range and acoustic environment of the
sample data can be reproduced. Here are some examples.
Show this thread
0:51
109.2K views
6
5
44
Show this thread
We should also acknowledge there are still chess grand masters. They didn't become obsolete. In fact, chess is more popular now than at any time in history.
Chess AI isn't an existential threat to human players. On the contrary, it helped make chess accessible to more people.
4
8
22
Show this thread
Every book is about to be an audiobook, with voice, pacing & language selection. Will narrators rebel?
Quote Tweet
Apple just released digital narration to help authors generate audiobooks using AI.
Listen below. The digital voices sound just like human narrators.
Show this thread
1:30
85.9K views
29
37
212
Show this thread
Apparently a lot do you don’t understand the risk of AI authoritarianism.
ChatGPT is pretty wholesome & moderate. This won’t be the case of an AI not based on social liberalism.
China developed AI racial profiling in 2019. Modern phrenology in the service of authoritarianism.
9
3
37
Show this thread
Because AGI is the new nuclear bomb.
It's an arm's race. If it can design new 20,000 medicines, it can also design 20,000 poisons.
We must get it first, based on social-liberal memes. We can't take the risk to let an AGI grow from authoritarian memes from China, Iran or Russia.
Quote Tweet
Replying to @timsoret and @csubagio
Who is this magical entity that’s coming for us?
Because such thing doesn’t exist yet. It’s a human made product, and we can absolutely decide the future we want here.
14
7
76
Show this thread
The loom was devastating to textile workers. But it eventually ensured that even the very poor were well clothed. On balance, an absolute blessing for humanity and only an absurdly short-sighted analysis could disagree.
4
5
23
It's coming for all of us.
The same conflict is appearing in science.
In panic, some are trying to ban AI assisted papers from scientific publications. As Yann Lecun explains here, this is short-sighted:
Quote Tweet
This thread exposes a basic misunderstanding.
Some believe that science evaluation comes down evaluating scientists' intrinsic abilities, skills, or merit.
For them, using AI tools is like "cheating".
But science must solely evaluate *impact*.
It's not a beauty contest. twitter.com/ylecun/status/…
1
4
27
Show this thread
Finally, an admission...
That the real issue with AI isn't copyrighted human data, but how it trivializes human effort overall.
That's a hard pill to swallow, for all of us. Now, it's a question of defending our selfish interests vs the common good. Choose your camp wisely.
Quote Tweet
Replying to @shadmbrooks
You say that its no different from someone editing your art in photoshop but are wrong yet again. You still need time and skilled input to make these kinds of changes in PS. The ease of use and scale at which these AI generators can modify images is the problem here.
51
15
160
Show this thread
Replying to
It's incredible.
There's a .pptx, and a SIGGRAPH video "Creating the Atmospheric World of Red Dead Redemption 2 " that's quite difficult to find, so I uploaded it here:
8
18
157