A.I. related stuff
category: code [glöplog]
Let me begin with a warning: this topic might cause some allergies. But for the hardened, - if you have any (entertaining) A.I. related articles, insight or videos, why not aggregate it here, as a collection to smile upon, from a future perspective, when most of us have lost their jobs. Haha.
Here's my personal quick-dive into the essentials:
-- An off-topic precursor to my interest in A.I. (2008)--
Many-Worlds Mario Kaizo level 1 (NOTE: NOT Artificial Intelligence! I know!)
-- A.I. shows up on the radar for me personally, as a usecase (2013)--
Computer program that learns to play classic NES games (Playfun)
-- Now you have my full attention (2016) --
Neural Network Learns to Generate Voice (RNN/LSTM)
-- Oh, okay.. (2017) --
But what *is* a Neural Network? | Deep learning, chapter 1
-- Wat... (2017) --
Example-Based Synthesis of Stylized Facial Animations
-- ... why... didn't i... (2016) --
TensorFlow in 5 Minutes (tutorial)
So. If you do not find this topic toxic and have exciting contributions like this (*joking*) or even real insight like articles and code examples, be brave and share them! Maybe someone can synthesize a demo out of it!
best,
WS
Here's my personal quick-dive into the essentials:
-- An off-topic precursor to my interest in A.I. (2008)--
Many-Worlds Mario Kaizo level 1 (NOTE: NOT Artificial Intelligence! I know!)
-- A.I. shows up on the radar for me personally, as a usecase (2013)--
Computer program that learns to play classic NES games (Playfun)
-- Now you have my full attention (2016) --
Neural Network Learns to Generate Voice (RNN/LSTM)
-- Oh, okay.. (2017) --
But what *is* a Neural Network? | Deep learning, chapter 1
-- Wat... (2017) --
Example-Based Synthesis of Stylized Facial Animations
-- ... why... didn't i... (2016) --
TensorFlow in 5 Minutes (tutorial)
So. If you do not find this topic toxic and have exciting contributions like this (*joking*) or even real insight like articles and code examples, be brave and share them! Maybe someone can synthesize a demo out of it!
best,
WS
Manufacturing an Artificial Intelligence Revolution (2017)
--
"While the term "Artificial Intelligence" (AI) was coined in the 1950s, in recent years AI has become a focus of attention in mainstream media. Yet the forces behind AI's revival have been unclear. I argue here that the "AI" label has been rebranded to promote a contested vision of world governance through big data. Major tech companies have played a key role in the rebranding, partly by hiring academics that work on big data (which has been effectively relabeled "AI") and helping to create the sense that super-human AI is imminent. However, I argue that the latest AI systems are premised on an old behaviorist view of intelligence that's far from encompassing human thought. In practice, the confusion around AI's capacities serves as a pretext for imposing more metrics upon human endeavors and advancing traditional neoliberal policies. The revived AI, like its predecessors, seeks intelligence with a "view from nowhere" (disregarding race, gender and class)---which can also be used to mask institutional power in visions of AI-based governance. Ultimately, AI's rebranding showcases how corporate interests can rapidly reconfigure academic fields. It also brings to light how a nebulous technical term (AI) may be exploited for political gain."
This list in general also seems like a good resource, though I haven't read most yet: https://github.com/chobeat/awesome-critical-tech-reading-list
--
"While the term "Artificial Intelligence" (AI) was coined in the 1950s, in recent years AI has become a focus of attention in mainstream media. Yet the forces behind AI's revival have been unclear. I argue here that the "AI" label has been rebranded to promote a contested vision of world governance through big data. Major tech companies have played a key role in the rebranding, partly by hiring academics that work on big data (which has been effectively relabeled "AI") and helping to create the sense that super-human AI is imminent. However, I argue that the latest AI systems are premised on an old behaviorist view of intelligence that's far from encompassing human thought. In practice, the confusion around AI's capacities serves as a pretext for imposing more metrics upon human endeavors and advancing traditional neoliberal policies. The revived AI, like its predecessors, seeks intelligence with a "view from nowhere" (disregarding race, gender and class)---which can also be used to mask institutional power in visions of AI-based governance. Ultimately, AI's rebranding showcases how corporate interests can rapidly reconfigure academic fields. It also brings to light how a nebulous technical term (AI) may be exploited for political gain."
This list in general also seems like a good resource, though I haven't read most yet: https://github.com/chobeat/awesome-critical-tech-reading-list
noby: that's a great list, thank you!
semi-related: i just recalled watching this movie, which alot of people hated, - i actually liked it for some reasons - Automata (2014). but i guess it was probably way too dystopic for most viewers.
semi-related: i just recalled watching this movie, which alot of people hated, - i actually liked it for some reasons - Automata (2014). but i guess it was probably way too dystopic for most viewers.
Shameless plug:
http://www.ai-healthcare.net/
http://www.21stcenturyheadlines.com
(I already know what statements from the latter website are going to trigger Maali & Co. now.... I am accustomed to that meanwhile :))
http://www.ai-healthcare.net/
http://www.21stcenturyheadlines.com
(I already know what statements from the latter website are going to trigger Maali & Co. now.... I am accustomed to that meanwhile :))
As in material quoted by Noby, AI is obviously extremely overused term nowadays.
It's puts everything related to data analysis, regression models, computer vision, search algorithms etc... into one big pile of "deep" bullcrap.
Nevertheless, strong AI is actually interesting.
In the past (early 2000s), I had few friends working on (in)famous ATR's Brain project led by Hugo de Garis. Unfortunately, the project failed *spectacularly* at the weight of its own bullshit - project leads raised investor expectations to infinity and beyond, but in reality they couldn't even deliver 1% of what they promised.
If anyone is interested, I have found some fairly recent survey of similar projects apparently ongoing written by nobody else but de Garis himself (yeah, 2010 is not exactly recent, but it's not so bad in the realistically measured progress in the field).
It's puts everything related to data analysis, regression models, computer vision, search algorithms etc... into one big pile of "deep" bullcrap.
Nevertheless, strong AI is actually interesting.
In the past (early 2000s), I had few friends working on (in)famous ATR's Brain project led by Hugo de Garis. Unfortunately, the project failed *spectacularly* at the weight of its own bullshit - project leads raised investor expectations to infinity and beyond, but in reality they couldn't even deliver 1% of what they promised.
If anyone is interested, I have found some fairly recent survey of similar projects apparently ongoing written by nobody else but de Garis himself (yeah, 2010 is not exactly recent, but it's not so bad in the realistically measured progress in the field).
tomkh -- thx!
i also wondered how the original idea of self-designing machines. turing-test surroundings and a huge bundle of philosophical and ethical problematics were suddenly just boiled down to a modernized approach of just concentrating on big data.
to me, any kind of marketing hype in that direction is okay, since i hope that someone resonates with the topic who really pushes it forward -- and perhaps unorthodox approaches bring unexpected results.
i am actually surprised, that there is not a much bigger hype.
i also wondered how the original idea of self-designing machines. turing-test surroundings and a huge bundle of philosophical and ethical problematics were suddenly just boiled down to a modernized approach of just concentrating on big data.
to me, any kind of marketing hype in that direction is okay, since i hope that someone resonates with the topic who really pushes it forward -- and perhaps unorthodox approaches bring unexpected results.
i am actually surprised, that there is not a much bigger hype.
I fail to grasp your last statement. IMHO, the hype is huge, indeed! I follow dozens of websites related to science and technology every day... AI is definitely the number one topic!
However, I personally am far more interested in things such as code optimization, Assembler programming etc. than in actually implementing backpropagation and similar stuff. That's why my own approach to AI is more of a theoretical nature. (However, I've done Andrew Ng's course on Coursera, and solved all programming exercises in it.)
IMHO, AI is something for a new generation of computer scientists, younger than me (I was born in 1983). It is quite different from what us demosceners are used to doing. You do not need the skills we acquired by coding demos. You need different skills. I would not say doing AI stuff is easier than coding demos, it is just different.
As a matter of fact, during my student days I specialized in evolutionary algorithms. That is an approach that is sometimes also considered related to AI. There are some publications that show that evolutionary algorithms are sometimes more effective than deep learning, e.g.
https://www.ascend.ai/de/blog/evolutionary-algorithms-the-next-big-thing-in-machine-learning/
https://www.sciencedirect.com/science/article/pii/S2210650217302766
https://www.analyticsvidhya.com/blog/2018/07/evolutionary-algorithm-perfect-useful-alternative-neural-network/
However, this is not at all related to my current job (which I have been doing for 4.5 years now), which is about computational physics...
However, I personally am far more interested in things such as code optimization, Assembler programming etc. than in actually implementing backpropagation and similar stuff. That's why my own approach to AI is more of a theoretical nature. (However, I've done Andrew Ng's course on Coursera, and solved all programming exercises in it.)
IMHO, AI is something for a new generation of computer scientists, younger than me (I was born in 1983). It is quite different from what us demosceners are used to doing. You do not need the skills we acquired by coding demos. You need different skills. I would not say doing AI stuff is easier than coding demos, it is just different.
As a matter of fact, during my student days I specialized in evolutionary algorithms. That is an approach that is sometimes also considered related to AI. There are some publications that show that evolutionary algorithms are sometimes more effective than deep learning, e.g.
https://www.ascend.ai/de/blog/evolutionary-algorithms-the-next-big-thing-in-machine-learning/
https://www.sciencedirect.com/science/article/pii/S2210650217302766
https://www.analyticsvidhya.com/blog/2018/07/evolutionary-algorithm-perfect-useful-alternative-neural-network/
However, this is not at all related to my current job (which I have been doing for 4.5 years now), which is about computational physics...
BTW, in the past year, I delivered two interviews related to AI, which you might be interested in, too:
https://goodmenproject.com/featured-content/volko-sjbn/
https://in-sightjournal.com/2018/06/01/rosner-volko-one/
https://in-sightjournal.com/2018/06/08/rosner-and-volko-two/
https://in-sightjournal.com/2018/06/15/claus-and-rosner-three/
https://in-sightjournal.com/2018/06/22/claus-and-rosner-four/
https://goodmenproject.com/featured-content/volko-sjbn/
https://in-sightjournal.com/2018/06/01/rosner-volko-one/
https://in-sightjournal.com/2018/06/08/rosner-and-volko-two/
https://in-sightjournal.com/2018/06/15/claus-and-rosner-three/
https://in-sightjournal.com/2018/06/22/claus-and-rosner-four/
Lesen Sie was ich sagte sich selbst doch versuchsweise laut vor. Vielleicht können Sie es besser verstehen, wenn die Worte aus ihrem eigenen Munde kommen. Ich finde dass der sich ständig wiederholende Hinweis auf die eigene Intelligenz einen fahlen Beigeschmack liefert, der die Frage in den Raum stellt, warum die eigenen Thesen nicht ausreichen, um auf den Umstand der hohen Intelligenz hinzuweisen. Zumal es dann eben andere Thesen aus der gleichen Quelle gibt, die eigentlich der hohen Intelligenz auf dramatische Weise zu widersprechen scheinen.
Vielen Dank jedoch für Ihren Beitrag.
Vielen Dank jedoch für Ihren Beitrag.
LOL
Videos related to AI from the World Economic Forum 2019 (January 2019, Davos, Switzerland):
Setting Rules for the AI Race
https://www.weforum.org/events/world-economic-forum-annual-meeting/sessions/setting-rules-for-the-ai-race
Compassion through Computation: Fighting Algorithmic Bias
https://www.weforum.org/events/world-economic-forum-annual-meeting/sessions/compassion-through-computation-fighting-algorithmic-bias
Automated Markets
https://www.weforum.org/events/world-economic-forum-annual-meeting/sessions/automated-markets
Setting Rules for the AI Race
https://www.weforum.org/events/world-economic-forum-annual-meeting/sessions/setting-rules-for-the-ai-race
Compassion through Computation: Fighting Algorithmic Bias
https://www.weforum.org/events/world-economic-forum-annual-meeting/sessions/compassion-through-computation-fighting-algorithmic-bias
Automated Markets
https://www.weforum.org/events/world-economic-forum-annual-meeting/sessions/automated-markets
Just a quick note: There is an excellent AI researcher named Daniel Sýkora who has his roots in the demoscene (his nickname was Redox). Just take a look at one of the many amazing things he is doing: https://www.youtube.com/watch?v=Ahft60eEUT8
found out the other day Sophia is coming to Cyprus, regular price tickets are €450
DeepMind recently announced a new bot named AlphaStar for playing StarCraft2. It can beat pro players.
https://www.youtube.com/watch?v=UuhECwm31dM
Robert Miles has made several good videos about AI Safety.
https://www.youtube.com/channel/UCLB7AzTwc6VFZrBsO2ucBMg/videos
https://www.youtube.com/watch?v=UuhECwm31dM
Robert Miles has made several good videos about AI Safety.
https://www.youtube.com/channel/UCLB7AzTwc6VFZrBsO2ucBMg/videos
afaik most of the original footage in this movie https://www.imdb.com/title/tt7905466/ was upscaled and enhanced using algorithms that are based upon deep learning, which is often cited to be some sort of artificial intelligence.
but since there is a d̶e̶m̶o̶ video about it, i might have been mislead earlier. https://www.youtube.com/watch?v=_cSXfKSRKz4
[related!] https://thoth.inrialpes.fr/src/deepflow/
Elon Musk on what his first question would be to a real A.G.I. system:
- What's outside of The Simulation?
https://www.youtube.com/watch?v=dEv99vxKjVI#t=1912
- What's outside of The Simulation?
https://www.youtube.com/watch?v=dEv99vxKjVI#t=1912
Just a small update on deep learning and faceswapping.
Few-Shot Adversarial Learning of Realistic Neural Talking Head Models https://www.youtube.com/watch?v=p1b5aiTrGzY
Paper: https://arxiv.org/abs/1905.08233v1
Few-Shot Adversarial Learning of Realistic Neural Talking Head Models https://www.youtube.com/watch?v=p1b5aiTrGzY
Paper: https://arxiv.org/abs/1905.08233v1
Adok I want to learn more about you and your brilliance! Enlighten our simple minds! Pleas post some more!
I really liked Andrej Karpathy classes https://www.youtube.com/playlist?list=PLkt2uSq6rBVctENoVBg1TpCC7OQi31AlC
His blog http://karpathy.github.io/
His blog http://karpathy.github.io/
Found this on LinkedIn:
Linear Algebra for Machine Learning (62 pages) - Sargur Srihari
Other Great Articles: (downloadable)
📌 Visual Intro to NumPy and Data Representation Python (14 pages)
https://lnkd.in/dpR2ez2
📌 41 Essential Machine Learning Interview Questions & Answers
https://lnkd.in/d8n2-Wp
📌 Machine Learning Projects with Python
https://lnkd.in/d6zZEw4
📌 Predicting the Future with Mark Cuban
https://lnkd.in/eYndTaR
📌 Top Data Science, Machine Learning Methods, Algorithms used
https://lnkd.in/denBHU4
📌 Walkthrough example of how to use Deep Learning to solve a problem
https://lnkd.in/dEf-6rV
📌 Machine Learning In The Real World
https://lnkd.in/dQsXvk4
📌 Comparison of Classification algorithms
https://lnkd.in/d6fNXYy
📌 An Introduction to Convolutional Neural Networks
https://lnkd.in/dnyqrVE
📌 The Mathematics of Machine Learning
https://lnkd.in/deiKZvG
📌 Google Machine Learning Dictionary
https://lnkd.in/dPieii6
📌 The GIT cheatsheet
https://lnkd.in/dcMcc9q
💙 The Ultimate Guide to AI, Data Science & Machine Learning
https://lnkd.in/eiPz3GJ
Linear Algebra for Machine Learning (62 pages) - Sargur Srihari
Other Great Articles: (downloadable)
📌 Visual Intro to NumPy and Data Representation Python (14 pages)
https://lnkd.in/dpR2ez2
📌 41 Essential Machine Learning Interview Questions & Answers
https://lnkd.in/d8n2-Wp
📌 Machine Learning Projects with Python
https://lnkd.in/d6zZEw4
📌 Predicting the Future with Mark Cuban
https://lnkd.in/eYndTaR
📌 Top Data Science, Machine Learning Methods, Algorithms used
https://lnkd.in/denBHU4
📌 Walkthrough example of how to use Deep Learning to solve a problem
https://lnkd.in/dEf-6rV
📌 Machine Learning In The Real World
https://lnkd.in/dQsXvk4
📌 Comparison of Classification algorithms
https://lnkd.in/d6fNXYy
📌 An Introduction to Convolutional Neural Networks
https://lnkd.in/dnyqrVE
📌 The Mathematics of Machine Learning
https://lnkd.in/deiKZvG
📌 Google Machine Learning Dictionary
https://lnkd.in/dPieii6
📌 The GIT cheatsheet
https://lnkd.in/dcMcc9q
💙 The Ultimate Guide to AI, Data Science & Machine Learning
https://lnkd.in/eiPz3GJ
https://www.youtube.com/watch?v=wWCiviGLVnk
Neural network generating technical death metal, via livestream 24/7 to infinity.
Neural network generating technical death metal, via livestream 24/7 to infinity.