Kozminski Techblog

Reading Time: 3 minutes

In 1 e. 2s of “Black Mirror” we observe rather usual situation for sci-fi series: 

Martha’s boyfriend Ash is killed while returning a hire van…Martha learns she is pregnant and service, by aggregating Ash’s many social media posts and online communications, an artificial intelligence (AI) imitation of Ash is created. Martha interacts with him via instant messaging and video call…The next stage is a physical android version of Ash, which begins to make Martha uncomfortable. She argues with him and takes him to a cliff where she orders him to jump, but Martha is then frustrated as the real Ash would not obey such a command. Several years later, on Martha’s daughter’s birthday, she takes a slice of cake to the android Ash kept in the attic, which she is only permitted to visit on weekends.

The idea of talking with deceased isn’t new: from the beginning of humanity, when when we were still living in tribes and using opium, cannabis, alcohol, tobacco, various forms of fungi to experienced enlightenment , throughout Salem witch trials and other religious persecutions, till notorious victorian spiritualism practices, and people still try to reach out to their deceased loved ones, but someone still uses spiritual desk and someone using an AI.

We all, probably understood, that black mirror’s episode plot may be real in some time, but most of us couldn’t think that somebody in only several years would really took a swing at , in some way, sacred and very intimately-emotional thing as death and artificial resurrection of the loved one. 

Isn’t it kinda a Babylon tower thing? 

But, in the same, isn’t the most vulnerable places of human nature can bring the biggest money, huh?

And, what’s more, this somebody is none other than Microsoft, so it means that probably in distant future the project will take big turnover, but the company isn’t planning to turn the technology into an actual product in near future, because, as Tim O’Brien, Microsoft’s general manager of AI programs, said “it’s predates the AI ethics reviews we do today.” Probably, they may be afraid of too strong negative reaction to this product, but should we blame them?

According to the patent information, the tool would cull “social data” (images, social media posts, messages, voice data and written letters) of the individual. That data would be used to train a chatbot to “converse and interact in the personality of the specific person.” It could also rely on outside data sources, if user ask a question that couldn’t be answered based on the person’s social data.

“Conversing in the personality of a specific person may include determining and/or using conversational attributes of the specific person, such as style, diction, tone, voice, intent, sentence/dialogue length and complexity, topic and consistency, as well as using behavioral attributes such as interests and opinions and demographic information such as age, gender and profession, the patent states”.

But the real question is would you use this technology in the future?

https://www.syfy.com/syfy-wire/microsoft-AI-chatbot-patent-talk-to-dead-people

https://www.washingtonpost.com/technology/2021/02/04/chat-bots-reincarnation-dead/

Reading Time: 2 minutes

Kanye West, or Ye, is known as a great musician, who has changed the game of the whole rap industry, and not even once.  

He is a rapper. He is a fashion icon. He is a revolutioniser. He is bipolar.

Is he an inventor now? 

 In august Ye released “Donda” album and at the same time teased the prototype of Stem Player, which was made in partnership with Kano Computing and which promises the ability to “customize any song” and “manipulate music in real-time”.

Sounds rather dimly and abstract, but you should take a look at this video. 

How does it work?

Basically you manipulate the “Stems” (different aspects of a track like the vocals or the drums) from songs off Ye’s album Donda. Or, thanks to machine learning, it can pull off the same trick with any* album you put on it. Tools include the ability to create loops, raise and lower pitch, and even combine stems from different songs.

*Almost any – review on Engadget says that it’s handled splitting other tracks it rather disappointing way: 

“The Stem Player handled “DIRTY!” from JPEGMAFIA and “Stonefruit” by Armand Hammer reasonably well, though there was some slight bleed through of the synth line into the vocal stem on “Stonefruit.” Nine Inch Nail’s “Closer” fared a little worse. Half the bass line was on its own track, while the other half was lumped in with the drums. It even bled into the vocals…”

Also, we can’t skip the design part of the device: 

Alex Klein, CEO of Kano, says that during creation process they asked themselves such questions as “How do we make technology that doesn’t feel like it’s from an evil alien world? How do we make technology that we just naturally feel that we can control?”, – then he adds – “Ye was always wise on this. He has a great line. He says, “Humans are God’s greatest technology.” So, when we humans make technology, it should feel more like an extension of our bodies. It should feel more natural and organic, but also decidedly human”.

At this point, I’m concerned that this gadget+concept has big potential. Firstly, because it may change the whole DJ’s game by giving them much more opportunities at mixing, sampling and etc.

Secondly because it gives us an opportunity to be the creator and consumer at the same time.  

It’s like music Lego and who doesn’t like Lego?

sources:

https://www.complex.com/music/kanye-west-stem-player-alex-klein-interview/

https://www.engadget.com/donda-stem-player-kanye-west-kano-hands-on-150038410.html

Reading Time: 3 minutes

Lately, we can hear about this «unknown animal” from everywhere, and usually it comes in hand with big money. 

“Grimes sold $6 million worth of digital art as NFTs” – the Verge 

“Jack Dorsey sells his first tweet as an NFT for $2.9 million” – CNBC news

‘Charlie Bit My Finger’ YouTube video sells as NFT for $760,999” – NBC News

“Quentin Tarantino To Release Unseen ‘Pulp Fiction’ Scenes As NFTs” – the Screencrush

And more.

What is this? Is this even real? For what people paying these amounts? Let’s dive into. 

Non-fungible tokens use blockchain technology to certify the authenticity and ownership of a specific and unique digital object. 

Non-fungible means irreplaceable, what also means it’s kinda untradable, because, for example, you can trade one bitcoin for another and you will have the same thing, so it’s tradable, but with NFTs it doesn’t work like that. Why? Because it’s art and in fact an evolution of art (probably it’s correct to say digital art*) collecting.  It implies that basically you can buy everything ever done in internet: videos, photos, tweets, posts, music, images, digital art, design, your brain downloaded and turned into an AI etc.

Basically at allows you to buy and sell the ownership – something that can’t be copied. You get to own a piece of digital work that no one else can possess*. It’s possible to prove who owns a given NFT at any moment in time and trace the history of prior ownership.   

*BUT digital assets of an NFT can be duplicated as easily as downloading a jpeg to your desktop. The point is the ownership which is carved in stone in the blockchain (though the artist can still retain the copyright and reproduction rights, just like with physical artwork). To put it in terms of physical art collecting: anyone can buy a Van Gogh print. But only one person can own the origina

Which NFTs are popular? 

One of the most successful collections in the nonfungible token space – BAYC (Bored Ape Yacht Club) launched in April and consists of 10,000 Bored Apes. The cheapest you can buy one for now is 49 ether = $231,000. Post Malone, Jimmy Fallon and Steph Curry are among today’s most prominent owners. Basically it’s like Pokémons for rich. Bored Ape Yacht Club was launched by a team of four pseudonymous developers: Gargamel, Gordon Goner, Emperor Tomato Ketchup and No Sass. It took 12 hours for all 10,000 to sell out at a price of 0.08 ether ~ $190. 

What makes BAYC or any other NFT collection valuable is questionable in the same way as questionable every price tag for physical art is. Value is in the eye of the beholder. 

What about NFT’s future? Is it going to thrive on, or, in the same way as it gained quite an overnight success, it’ll loose it’s popularity in the in a heartbeat?

In my personal opinion it will thrive and prosper, because if we take a look into the history, we see that none of art forms has disappeared. None. Art is the most irrational process and obviously we could have managed without painting our clay dishes, but we still do this to this day. NFT is one of the new forms of art, so in the same way as cinema, music, pottery etc didn’t die out, NFT won’t die out.

sources:

https://www.marca.com/en/lifestyle/2021/12/01/61a762abe2704e455d8b4571.html

https://www.livemint.com/brand-stories/what-are-nfts-11636533998290.html

Reading Time: 2 minutes

Find the name which doesn’t fit the row: 

Johannes Brahms, Ludwig van Beethoven, Johann Sebastian Bach, AI, Pyotr Ilyich Tchaikovsky.  

Yes, you are right: Pyotr Ilyich Tchaikovsky, he’s russian composer, not german one.

How do you think, this “joke” in the near future won’t be joke anymore or are we overestimating the power of the AI which is, slowly, but confidently, on the way of taking over some job industries? 

A group of scientists, who worked with an AI start-up Playform AI have an intricate thing to come up with: now we are able to listen to the unfinished symphony of the music composer, who died 194 years ago. 

Ludwig van Beethoven left 9 symphonies and yet one completely unfinished piece: the 10th symphony was only some raw sketches on different sheets.

The team consisted of Austrian composer Walter Werzowa, Mark Gotham, a computational music expert, Robert Levin, musicologist at Harvard University. Why do we need this kind of crew?

In fact, music is very similar to language. That’s why natural language processing model is very close to what was done in this case. In the same way as when you allow the predictive text feature to end one word you’ve started typing or predict the next one – it usually works, but several words after it starts writing complete gibberish – the same case here. 

The problem here was that Ai couldn’t manage to continue an uncompleted piece of music beyond a few seconds.

But in the same way music uses mathematical structure but the language doesn’t. 

They used everything what Beethoven hypothetically heard growing up, because obviously there is much more examples of texting than Beethoven’s music. 

What’s interesting, is that they’ve tried to take in count the fact composer’s complete loss of hearing in the end of the life, because it’s influenced his music in the way of gradually adding more lower notes to his compositions, that he could hear more clearly.

But in the end of the day the symphony is written, and you can listen to it .

 Open question number one: should Beethoven, who in this case wrote less than 5% , be in credits as an actual author? Is it even should be considered as one of his compositions?

Open question number two:

What does it mean for the future of music?

Werzowa says about it in podcast: “Like every tool, you can use a knife to kill somebody or to save somebody’s life, like with a scalpel in a surgery. So, it can go any way … I think if we could sit back on a Saturday afternoon in our kitchen, and because maybe we’re a little bit scared to make mistakes, ask the AI to help us to write us a sonata, song or whatever in teamwork, life will be so much more beautiful”.

Reading Time: 2 minutes

How many of us, students, experienced problems with falling asleep, keeping concentration, trying to be focused? What about ADHD or inability to meditate because our “eternal sunshine of the spotless mind” can’t keep on without thinking every 15 minutes what do we have to buy at the grocery store or why does your back hurt so much or… etc.  Probably, at least one of these problems was faced by 90% of us , if not 100%. 

Now let me introduce the Muse – the most widely used consumer neurotechnology in the market today.

Muse is a brain fitness tool that measures brain signals much like a heart rate monitor senses your heartbeat. Muse’s 7 sensors, 2 on the forehead, 2 behind the ears plus 3 reference sensors, detect and measure the activity of your brain. This technology does not use electrical stimulation – it’s a passive tool that gives you accurate, real time feedback on what’s happening in your brain. 

Signal processing and machine learning techniques are applied to the brain signal components to control the experience, as well as dry sensor technology,  buetooth, digital signal processing make it easy to access and use brainwave data.

When your mind is calm and settled, you hear peaceful weather. Once your brain starts thinking about things unconsciously, you’ll start to recognize that the ocean wave sounds start crashing louder and the wind sounds start picking up. 

Basically it’s electroencephalography (EEG) system in your pocket. 

Neurofeedback began gaining popularity years ago in clinical contexts, as research showed (https://www.washingtonpost.com/national/health-science/therapists-are-using-neurofeedback-to-treat-adhd-ptsd-and-other-conditions/2015/01/16/b38e6cee-5ec3-11e4-91f7-5d89b5e8c251_story.html) it had the potential to help people struggling with conditions like ADHD and PTSD.

In conclusion, I believe that technologies+neuroscience/mental health field is the most promising combination which will thrive and prosper in future, because health (both: physical and mental) is always one of the most vital topics ever and technologies allow us to improve and upgrade understanding of how human’s brain and body work, as well as to help us to cope with diseases.

sources: https://www.vox.com/future-perfect/22727109/enlightenment-technology-neurofeedback-brain-stimulation-psychedelics