Capitalogix

Another AI Art Gimmick

There's a new trend of AI-generated profile pictures - using tools like avatar.ai.

Below are some sample results my son, Zach, got after downloading an iPhone app called Lensa.  The company that makes it also offers a product called Prisma that is worth looking at as well.  He paid $4 within Lensa to generate these AI photos (for context, avatar.ai charges more than 5x as much).  He uploaded 17 photos from different angles (and with different expressions), and the app promised 50 pieces from 4 styles.

What do you think of the results?  I thought some of them were surprisingly good. 

 

Some were not so good ... (or at least outside my zone of artistic preferences).

After seeing his, I'll admit I got curious and made my own. 

 

To be honest, these apps are still volatile ... meaning, sometimes they are good enough to seriously impress – but they also can underwhelm from time to time.  Check out DALL-E for more examples of the dynamic range of possibilities and results.

Regardless, I am impressed with AI art's progress and momentum.  Recently, generative AI has gotten a lot of hate (along with the hype), nonetheless, people are talking about it (meanwhile AI art is still not as polarizing as Kanye). 

The reality is, most of these tools still are an elegant use of brute force.  Nonetheless, I am bullish.  Hardware and software are getting better ... and when fed more data, generative art depictions get better as well. 

The uses continue to get more elegant and complex as time passes ... but we're still coding the elegance. 

For more on this topic: 

• Dall-E and AI-Generated Art 
• AI editing underwater photography
• AI evaluates how normal your face is
• On deepfakes and fake news
• Bruce Willis and AI movie characters

The Swish Machine

Rube Goldberg machines impress me more often than not. Here's a 70-step outdoor machine that covers a lot of ground ... all to put a basketball in a hoop. It took a month to create, and another month to get working. 

 

via Creezy

I love stuff like this because it reminds me of life. Looking backward, you see how all the pieces go together. As you're going through it, it feels random and sometimes like you're moving backward or that your effort isn't directly contributing to your goal.

Nonetheless, it all comes together in the end.

Onwards!  

"Earl, Honey"

 

My first wife (and mother of my children) Denise is a smart and talented author.  I still remember the first time she asked me to read an early copy of her work.  She never let me do it again. 

That being said, she recently released a new book, Earl, Honey.  The book is loosely based on the real story of her family history.  I enjoyed it.  It's a southern coming-of-age story set in the 1920s - which is not my typical genre.  Nevertheless, it's a poignant story with insight into the human condition, and its dramatic realities remind us how good we have it. 

It's a story she learned, first via her grandmother, long after the events of the book take place.  Those events shaped the lives of generations of her family. 

It's a tough and heart-wrenching Southern Gothic read that covers incest, domestic abuse, and more.  Check it out. 

Here's the book's blurb from Amazon:

"Ever since Pa hit him in the head with the two-by-four, Earl had lived with blinders on. Not real blinders, of course, because that would be foolish. It was his own brain that blinkered him."

Earl Hahn is slow, the last one to catch on to things. Since the day his father hit him in the head with a 2x4 of loblolly pine, he's struggled with a "thickness in his brain." It takes him longer to make the connections others arrive at easily. When his father is prosecuted for the crime of incest, it feels like deliverance for Earl, his mother Lizzie Belle, and the entire Hahn family. Unfortunately, his father's abhorrent actions are not done exacting a price.

Everyone in the household will pay for their patriarch's crimes - no one more than Earl.

So begins a powerful coming-of-age tale about a shy, damaged boy who must overcome unimaginable personal tragedy - both as its victim and its perpetrator. Raw, honest, and filled with heart, Earl, Honey recounts an extraordinary search for redemption amid the perilous world of the 1920s American South.

 

A Wordcloud For Each of the Major Religions

The six largest religions in the world are Christianity, Islam, Judaism, Hinduism, Buddhism and Sikhism. 

We often think about the differences between religions.  However, the many similarities are obvious if you look (and may be indications of a more integral "truth". 

Below is a wordcloud for each of those religions based on their major religious text. 

teddyterminal via Reddit

If you find the name "Keith" it's because it was the translator's name, and the word "car" in the Hinduism wordcloud is an old-fashioned word for "chariot".

It's also worth acknowledging that this wordcloud is from the English translations so some words that may mean slightly different things in other languages can be all translated to one word in English. For example, it's very common in Biblical Hebrew to see different words translated into the same English word. A good example is Khata, Avon, and Pesha--three different ways of committing a wrong, that may all be translated the same.

AI Meets Dr. Seuss

Dr. Seuss was recently in the news for stopping the release of 6 of its books. 

Whether it was a marketing ploy or not, I've been seeing a lot more Dr.Seuss content. 

To start, here's a video of an A.I. written Dr. Seuss book with animation. 

 

via Calamity AI

In addition, here's an A.I. remastered World War II cartoon written by Dr. Seuss with a character named Private Snafu. It's one episode of a series of shorts that were banned post-WWII, and it's one of the more tame episodes.  For an extra piece of trivia, the name of Private Snafu and his series of shorts was based on the military acronym for "Situation Normal: All F***ed Up". 

It's an interesting piece of history ... enjoy. 

 

via Adam Maciaszek

While produced by Warner Bros., these shorts which were made for the US military did not have to go through the Production Code Administration and thus got away with raunchier humor, foul language, and what we would today categorize as racist propaganda against the Japanese and Germans. 

While it's okay to acknowledge that we should be doing better today, I also think it's interesting and informative to watch older materials in the context and time period they were written. 

Racism isn't okay, but if you don't know history, you're doomed to repeat it, and art can be discussed and enjoyed within that context as well.  

 

Data Really Is Beautiful

I think most data scientists or traders would agree that some charts are just prettier than others.

Whether it's due to the artistry of the creator, the results shown, or an insight or perspective illuminated ... I am sometimes surprised by the beauty of a chart. 

After looking at thousands of charts, some really do look "pretty" and others look "ugly" to the trader.  Perhaps this stems from an intuition honed through many trials of separating luck from skill?

Taking a different approach is Stoxart, created by a visual designer at Nike named Gladys Orteza.  She has been turning stock charts into landscape artworks related to the company they reference.  All that's missing is the warning that past performance doesn't guarantee future results.

Here is an example of her art inspired by Ford's performance in the last year.  Maybe she should have titled it "Sunset".

via LLMoonJ

 Another fun one is a year of Tesla performance. 

via LLMoonJ

Here's a link to see more Stoxart. 

Mummy, I've Lost My Voice

Some applications of technology are just ... strange. That being said, I think there's something beautiful about science being done "just to see if we can". 

Not every experiment needs to be in pursuit of some grand truth or miracle cure... Sometimes it's nice just to be curious. 

In that regard, the researchers who recreated a 3,000-Year-Old Egyptian mummy's voice might "take the cake" (or "drop the mic"). 

 

Nature via Leeds Teaching Hospitals (Click To Learn More)

The mummy's name is Nesyamun. They used a 3D printer and an electronic larynx to create the sound. They didn't recreate his tongue - so his voice doesn't take that into account. From what I can tell, all they've gotten "him" to say is "ehh". Very mummy-like. 

Here's a link to the sound. 

For what it's worth, they did test this methodology on a living human to validate it before going to some 3000-year-old dead guy.

Here's an interview with one of the scientists on BBC's Inside Science. 

Here Are Some Links For Your Weekly Reading - January 19th, 2020

Here are some of the posts that caught my eye recently. Hope you find something interesting.

Lighter Links:

• Artificial Intelligence: How to Measure the I' in AI. (TheNextWeb)
• Samsung Sheds Light on Its 'Artificial Human' Project. (EnGadget)
• Secrets from a Former FBI Negotiator on How to Get What You Want. (TheLadders)
• Here's the Advice Mark Cuban Would Give His 20-Year-old Self. (CNBC)
• Want to Grow 10X? Here are 17 Books about Business and Life You Should Read in 2020. (TheLadders)

 

Trading Links:

• The Decade of the "Young Old" Begins. (WorldIn)
• 2019 A Year to Forget for Hedge Funds. (Benzinga)
• Fortnite Made $1.8 Billion in 2019. (CNN)
• Amazon Partners with Lamborghini, Rivian to Offer Alexa Voice Assistant in Vehicles. (CNBC)
• Apple Made Investors Half a Trillion Dollars Last Year. Time to Hedge Your Bets. (Barrons)

Sea-Thru Sea

An engineer and oceanographer, Derya Akkaynak, created an algorithm that removes the water from underwater images - meaning it takes away the haze & tint that come with most underwater photos. It doesn't require a color chart either (though it does need distance information it gathers through numerous photographs from different angles). 

Check it out. 

 

via ScientificAmerican 

From the abstract: 

The Sea-thru method estimates backscatter using the dark pixels and their known range information. Then, it uses an estimate of the spatially varying illuminant to obtain the range-dependent attenuation coefficient. Using more than 1,100 images from two optically different water bodies, which we make available, we show that our method with the revised model outperforms those using the atmospheric model. Consistent removal of water will open up large underwater datasets to powerful computer vision and machine learning algorithms, creating exciting opportunities for the future of underwater exploration and conservation.

Essentially, the algorithm goes through every pixel and color balances the image based on collected data on distance/color degradation.

As with most algorithms of this type - the more data we feed it, the better it gets. Sea-thru already has great applications for furthering ocean-based research, but the follow-up question is can this algorithm be extrapolated outward to deal with other atmospheric conditions outside of water - smog, etc. 

What other uses can you imagine?  It is not hard to imagine how this could be applied to market data either ... Interesting stuff! 

Happy Halloween: The Trick To Treats

Good data visualizations are few and far between. Bad data visualizations spread lies, misrepresent content, or are sometimes just plain unreadable. 

In honor of Halloween, here's an example of a fun data visualization put together by Mariona Banyeres based on FiveThirtyEight's "The Ultimate Halloween candy power ranking." Compared to simpler charts it may be harder to read, but it's interactive, it's themed, and if you hover over points, it will add extra context. Click to go to Mariona's interactive Tableau DataViz. 

Mariona Banyeres via FiveThirtyEight

 

To skip the data visualization, here are the results of FiveThirtyEight's competition. 

via FiveThirtyEight

Hope you have a Happy Halloween.