The "thing" next to the fingernail is a functioning computer created at the University of Michigan (and it measures just .3 mm a side). It's run by photovoltaics and is primarily used as a precision temperature sensor. The caveat is that once it loses power, it loses all prior programming and data.
It tests the limits of what we call a computer – but it's multitudes better than the previous iteration, and innovation breeds innovation.
AI has plenty of weaknesses – I've talked about some before, and I'll continue to talk about them in the future, but two specific weaknesses were brought to my attention this week.
AI Portraits – Won't Steal Your Data, But Might Steal Your Soul Dorian Gray-Style
I assume most of you have seen the FaceApp trend – people age-ifying their photos and unwittingly giving the rights to their photos to a shadowy Russian tech company. You've also likely seen AI paintings selling for ridiculous money.
But have you seen their lovechild AI Portraits – a more wholesome experiment run by the MIT-IBM Watson AI Lab. AI Portraits uses approximately 45,000 different Renaissance-esque 15th-century portraits and General Adversarial Networks to translate your selfie into an artistic masterpiece. It's novel because instead of simply drawing over your face it's generating new features and creating an entirely new version of your face.
Mauro Martino via YouTube
It's impressive because it determines the best style for your portrait based on your features, your background, and more.
However, it's not without "flaw". The choice of 15th-century portraiture creates a couple of clear biases. At the time, portraits of smiling or laughing individuals were rare, so your smile will likely not transfer. As well, there's a clear bias towards anglo-saxonification.
My son got excited while playing with the app and sent several of his coworkers, friends, and family through the app. If you look at the bottom right, you'll see my lovely wife Jen's portrait.
Most of you have seen my wife and know that she is Indonesian, something that is very much removed from the translation.
All photos are immediately deleted from their servers after creating your image, so your privacy is safe (this time!)
All biases can be considered quirks of this current iteration of the program – which I do earnestly believe is interesting.
Later, you can imagine an AI choosing between various different styles of art based on a cornicopia of factors – or off human selection – but you have to walk before you can run, and this is a fun way to get people excited about AI.
Computer Answering Systems – No, The Answer Isn't 42
“Yes…Life, the Universe, and Everything. There is an answer. But I’ll have to think about it…the program will take me seven-and-a-half million years to run.” – Deep Thought, Hitchhiker's Guide To The Galaxy
Think of the global excitement when IBM's Watson first beat Ken Jennings in Jeopardy … it's widely considered one of the holy grails of AI research to create a machine that truly understands the nuances of language and human thought. Yet, if you've talked to Alexa recently, you know there's a long way to go.
Today's question answering systems are basically glorified document retrieval systems. They scan text for related words and send you the most relevant options. Researchers at the University of Maryland recently figured out how to easily create questions that stump AI (without being paradoxical, impossible to answer, requiring empathy etc.) in order to enhance those systems.
A system that understands those questions will be a massive step toward a real understanding and processing of language.
So what's the secret to these "impossible" questions?
The questions revealed six different language phenomena that consistently stump computers. These six phenomena fall into two categories. In the first category are linguistic phenomena: paraphrasing (such as saying “leap from a precipice” instead of “jump from a cliff”), distracting language or unexpected contexts (such as a reference to a political figure appearing in a clue about something unrelated to politics). The second category includes reasoning skills: clues that require logic and calculation, mental triangulation of elements in a question, or putting together multiple steps to form a conclusion […]
For example, if the author writes “What composer's Variations on a Theme by Haydn was inspired by Karl Ferdinand Pohl?” and the system correctly answers “Johannes Brahms,” the interface highlights the words “Ferdinand Pohl” to show that this phrase led it to the answer. Using that information, the author can edit the question to make it more difficult for the computer without altering the question’s meaning. In this example, the author replaced the name of the man who inspired Brahms, “Karl Ferdinand Pohl,” with a description of his job, “the archivist of the Vienna Musikverein,” and the computer was unable to answer correctly. However, expert human quiz game players could still easily answer the edited question correctly.
The main change is increasing the complexity of the questions by nestling another question. In the above example, the second question forces the AI not only to decide the composer inspired by Karl Ferdinand Pohl, but also to decipher who is inspiring (hint: It's Karl Ferdinand Pohl).
AI isn't great yet at mental triangulation; at putting together multiple steps to form a conclusion. While AI is great at brute force applications – we're still coding the elegance.
A staggering 90% of all the world’s data (2.5 quintillion bytes per day) has been created in the past two years alone … and its value is rapidly rising.
Data is today’s “wild west” and the battlefield of today’s tech titans. Alphabet, Amazon, Apple, Facebook, and Microsoft all have an unprecedented amount of data (and power).
Rapid growth means little time to create adequate rules. Everyone’s jumping to own more data than the next and to protect their own data from prying eyes.
I see it in trading, but it’s pervasive in every industry and in our personal lives.
Having basic data and basic analytics used to be enough, but the game is changing. Traders used to focus on price data, but now you’re seeing an influx of firms using alternative data sets to find an edge. If you’re using the same data sources as your competitors and competing on the same set of beliefs, it’s hard to find a sustainable edge. Understanding the game they’re playing, and their rules are important, but that’s table stakes.
Figuring out where you can find extra insight, or where you can make the invisible visible, creates a moat between you and your competition, and it lets you play your own game.
I shot a video where I talk high-level about Data as fuel for your business. Check it out.
It is interesting to think about what’s driving the new world (of trading, of technology, of AI, etc.) and that often involves identifying what drove the old world. History has a way of repeating itself.
Before e-mails, fax machines were amazing. Before cars, you were really happy with a horse and buggy.
It’s in these comparisons that I think we can help explain the importance of data in today’s new world economics.
Petroleum has played a pivotal role in human advancement since the industrial revolution; it fueled (and still fuels) our creativity, technology advancements, and a variety of derivative products. There are direct competitors to fossil fuels that are gaining steam, but I think it’s more interesting to compare petroleum to data due to their parallels in effect on innovation.
The process of pumping crude oil out of the ground and transforming it into a finished product is far from simple, but anyone can understand the process at a high-level. You have to locate a reservoir, drill, capture the resource, and then refine it to the desired product – heating oil, gasoline, asphalt, plastics, etc.
The same is true for data.
You've got to figure out what data you might have, how it might be useful, you have to figure out how to refine it, clean it, fix it, curate it, transform it into something useful, and then how to deliver it to the people that need it in their business. And even if you've done this, you then have to make people aware that it's there, that it's changing, or how they might use it. For people who do it well, it's an incredible edge. – Howard Getson
Data can be seen as the fuel to the information economy and oil to the industrial economy. The amount of power someone has can be correlated to their control of and access to these resources … and, leaking of these resources can lead to extreme consequences.
Why Data Is Better Than Oil
The analogy works, but it’s just that, an analogy, and the more you analyze it, the more it falls apart. Unlike the finite resource that is oil, data is all around us and increasing at an exponential rate, so the game is a little different:
Data is – ultimately – a renewable resource. It’s durable, it’s reusable, and it’s being produced faster than we can process it.
Because it’s not a scarce resource, there’s no urge to hoard it – you can use it, share it, transform it, infinitely knowing that it won’t diminish. Data is more useful the more you use it.
As the world’s oil reserves dwindle, and renewable resources grow in popularity and effectiveness, the relative value of oil drops. It’s unlikely that will happen to data.
Also, while data transport is important, it’s not expensive the way oil is. It can be transported and replicated at light speed.
Using alternative data gives traders an advantage, but it doesn’t always have to be confidential or hard to find information. Traders now have access to vast amounts of structured and unstructured data. An important source that many overlook is the data produced through their own process or the metadata from their own trades or transactions.
In the very near future, I expect these systems to be able to go out and search for different sources of information. It's almost like the algorithm becomes an omnivore. Instead of simply looking at market data or transactional data, or even metadata, it starts to look for connections or feedback loops that are profitable in sources of data that the human would never have thought of. – Howard Getson
In a word of caution, there are two common mistakes people make when making data-driven decisions. First, people often end up slaves to the data, losing focus on the bigger picture. Second, even the most insightful data can’t predict black swans. It’s important to exercise caution.
The future of data is bright, but it’s also littered with potential challenges. Privacy concerns and misuse of data have been hot button topics, as have fake news and the ability of systems to generate misleading data. In addition, as we gain access to more data, our ability to separate signal from noise becomes more important.
The question becomes, how do you capitalize on data, without becoming a victim to it?
Baseball signals are purposely confusing to keep them a secret. Attempting to decode your opponents' signals is part of the game. As it turns out, machine learning is pretty good at figuring them out.
Mark Rober, a former NASA engineer on the Curiosity Rover, put together a good video showing its efficiency. It's an interesting application and a pretty good introduction to machine learning. It did the rounds in my office, and we liked it.
