Writer Tara Calishain shares tips and tools so you can learn to do cool stuff with technology.
Headlines about artificial intelligence are almost uniformly grim: it’s the worst thing ever, the robots are coming, they’re taking our jobs, Daleks will rule the earth. Behind the headlines, though, things are a bit more prosaic: AI has a long road ahead of it. While it’s good at some things, it’s awful at others, and I don’t take seriously any robots that have to levitate to get up stairs (sorry, Daleks).
If you want to get a sample of what AI can do right now, take a look at photo apps on the web. Editing photos is often daunting, but web apps make it easy. If you can upload a photo, you can use a web app to colorize it, remove the background, and more!
These tools can’t take the place of a full-fledged photo editor like GIMP or Photoshop, or even a heavy-duty web app like PicMonkey. On the other hand, they’re free, and there’s no learning curve. If you’re interested in editing photos but don’t know how to get started, these apps can help you dip your toe in the water and get a sense of the possibilities.
For comparison I’ll be using this photo of my great-great-grandmother, Munsie, and showing you what four different sites can do to enhance this picture.
Colorizing photos is equally an art and a skill, and an AI-based app is not going to come close to the ability of a human colorist. At the same time, not everyone has the time or the ability to learn to colorize. These apps won’t be perfect, but you’ll be surprised how good they are. Note also that not all coloring apps are equal. The two I’m going to show you here will give you very different results. Pick the one you like best or try a lot of different photos for comparison.
With Algorithmia, you can upload a picture or choose one from the web. The instant you do, Algorithmia will start colorizing it and provide you with a comparison image. You can download that or download the fully-colorized photo. Here’s a comparison of Munsie.
As you can see, the image is almost sepia-toned except for the red in the jacket. I suspect it’s because this picture was taken outside in full sunlight and the image is accounting for the glare.
Colourise treats the image quite differently. This app was developed by GovTech Singapore, and the images it used to train its AI to color photos was completely different, which shows up in how it colors the picture of Munsie:
As you can see, Colourise went more for a green tint than red. I like this better because of the enhanced difference between Munsie’s skin and the wall behind her. Furthermore, her hair looks more natural, and not like there’s been a sheen of color applied to it. All photos are different, though, and I wouldn’t settle on using Colourise exclusively based on the processing of one image.
Take Out the Background with Remove.bg
I’m willing to bet that you have a picture of someone where you’d like to remove the background. Whether it’s because you want to get rid of somebody behind you or because the setting isn’t flattering, removing background images can be so useful but at the same time so difficult. But not with Remove.bg. Try this tool with a photo and you will feel like a wizard. In just a few seconds you’ll get a clean image with no or very little background.
There are a couple of caveats to be aware of: Remove.bg works best for single individuals, and it only works on humans (can’t take the background out of your dog picture!). All you have to do is upload your photo. Here’s the picture of Munsie with the background removed:
It’s almost perfect in only seconds. The only thing Remove.bg missed is the arm beside Munsie (probably because the arm and her jacket are so close together) and a bit of shadow.
Remove.bg does offer the option of editing the image further before downloading it, but I found the tools a little awkward. If you’d be comfortable using that, you’d probably be more comfortable using GIMP or Photoshop, which would give you more options for cleaning up the image. (To be honest I’d probably just crop the photo a bit.)
Get Stylin’ with DeepArt
Maybe you want to glam up your picture a little bit and give it a little pizazz. DeepArt is here for you. Upload an image and choose a style, and it’ll apply the style to your image for free — with one condition.
The one condition is that as of this writing the queue for getting your image processed is over five days long. If you’re willing to pay €1.99 (about $2.25) you’ll get your image in fifteen minutes. I found myself pleasantly surprised by how Munsie’s image worked out:
Editing images doesn’t have to be complicated, difficult, or costly! And we have AI to thank for it. I suspect that we’re only going to see more of these apps over time. I’ll be able to do even more of this instant editing, and see my picture of Munsie in more new ways.
Featured image: Photo courtesy Tara Calishain
Editor’s note: The following imagined account of life in the near future will, to some, have the ring of science fiction. But such leading thinkers as inventor Ray Kurzweil, Carnegie Mellon robotics professor Hans Moravec, and Google co-founder Larry Page have been making predictions since the 1990s about what the world will be like when computing power exceeds the capacity of the human brain. This inflection point, which some anticipate as the most important human development since spoken language, has come to be known in the scientific community as The Singularity.
Ahhh, to be young again! Well … not really. The year is 2057, and I just turned 100. But I have the energy, strength, and looks of a 40-year-old. I’m running, playing tennis and golf just as I did decades ago. All those annoying aches and pains that were my constant companions starting in my 60s? Gone! I’ll explain my good fortune a bit later, but first let me explain how we all got here.
Predictions for the timing of The Singularity ranged from 2030 to 2060, though historians will note that it was actually reached in 2045.
As artificial intelligence (AI) got better and better, there was still debate in the scientific community as to whether computers could ever be self-aware. Many argued that machines couldn’t, because they don’t strive for what they desire. Well, now that The Singularity has passed, early experiments indicate that we may indeed be able to create machines with consciousness. Regardless, life has been changing dramatically. Here’s a snapshot of my own experiences.
I have a companion I call Oscar. Back when I was a youngster of 50 something, I owned a popular device known as the iPhone, which came with a crude personal digital assistant called Siri. Oscar is what’s known as an intelligent digital assistant, or IDA. He’s a distant cousin of Siri, but he operates on an entirely different level. Have you ever wondered what it’s like to have a genie do your bidding? That’s what living with Oscar is like. He reads my thoughts — well, just the ones I want him to hear — and he speaks to me via a microscopic device in my ear. I think of him as a person, not a thing, because for all intents and purposes he is nearly human.
Just the other day, I was taking in the Bleimann exhibit at the Metropolitan Museum of Art, and ran into a woman I had met at a conference a few years ago, but whose name I couldn’t remember. “Jessica,” whispered Oscar, as she approached, smiling. “There’s an 87 percent chance that she finds you attractive and would welcome the opportunity to get to know you better.” As we chatted over lattes, Oscar quickly correlated several data points from her profile that were available on the Web, and ascertained with a high degree of certainty her likes and dislikes. My personal Cyrano prompted me with the conversation topics that were points of compatibility.
As we parted (with plans for dinner arranged) my car was waiting for me as I scampered out of the coffee shop. Take me home, I thought, and silently off we went.
I’m old enough to remember when cars were driven manually, resulting in over 30,000 highway deaths a year in the U.S. Even when self-driving cars became commonplace decades ago, libertarians insisted on continuing to drive themselves, extending the legacy of bloodshed. Once human drivers were finally abolished, highway deaths and traffic jams became things of the past.
On the ride, Oscar briefed me on my work for the following day, set up a slew of virtual meetings, and queued up everything I needed to know in preparation. Intelligent machines perform most job functions these days, leaving only the creative and interpersonal tasks to us humans for whom it still makes sense to work.
Longevity technology became widely available about 20 years ago. The process involves microbots, tiny robots as small as a blood cell, coursing through your bloodstream, cleaning up the debris associated with aging, zapping cancer cells long before they cause symptoms, and removing plaque from arteries just as it forms.
It was a bizarre feeling getting younger as I got older. I had the distinct sensation of time going backwards. Imagine feeling just a little bit more youthful each waking day. This can’t go on forever, of course. My expected lifespan is about 150 years, according to Oscar. You see, even with rejuvenation, eventually the organs will wear out.
Meanwhile, having the extra years makes you infinitely more relaxed about your life. My 5th wife and I recently decided to split up, but we’re still great friends. After 25 years of marriage, we both wanted to explore other interests. Twenty-five to 40 years is the norm for a marriage contract these days — which corresponds roughly to the natural length of marriage in olden times when it was bookended by puberty and death.
In the 2050s, “amicable divorce” is not an oxymoron. With Oscar to help me, I’m never lonely. I think back not forlornly but fondly on my shyness when I was young — a testament to the geekiness that accompanied my technical acumen. Today, I’m no longer a techie. In fact, I’m on my third career. After what used to be a lifetime in the software field, I fulfilled my lifelong ambition to be a musician. That was a rewarding experience, but after a few decades it had run its course. Now, I’m a writer.
Reinvention became a buzzword at the turn of the millennium when the baby boomers suddenly found themselves with decades of “bonus years” following retirement from the workforce. The phrase has even more meaning now, when one is physically hale enough to restart one’s life, study a new field of endeavor, and apply oneself to it. And, of course, there’s plenty of time. As for me, I may move on from writing at some point, but I’ll never go back to a technical occupation — the machines these days design software better than I ever could.
As I mentioned, futurists have predicted The Singularity for decades. At the dawn of the computer age, Intel co-founder Gordon Moore observed that computing power doubled roughly every two years, a phenomenon that became known as Moore’s Law. Decades later, Moore’s Law remained constant and the doubling even speeded up a little. It was a simple matter to extend the lines and determine when computers would match the power of the human brain.
The Singularity represents a confluence of technologies. Brain scanning technology, for example, allows us to communicate with our devices (and occasionally other people), without having to mumble at our devices or touch them. It took a little practice to train my brain to communicate without revealing my deepest, darkest thoughts. Just as you can train specific muscles to perform tasks, you can teach your brain to send instructions while keeping those private thoughts walled off. That technology will eventually allow humans to transfer their brains to a computer, one neuron at a time. In effect, we will soon be able to reverse-engineer the brain, by scanning it and transferring all neural states to a computer, which would then contain the consciousness of the original person, with all memories up to that point intact. There are still some bugs to be worked out; but the potential to live forever as a conscious being, albeit without a body, is now within the realm of possibility.
It wasn’t an easy road to get to The Singularity, and not everyone was as fortunate as I was. Things really took off in the late teens and early ’20s as the amount of data produced exploded. With the growth of social media, people were increasingly willing to sacrifice their privacy for the sake of convenience and connectivity. Since there was so much data out there, it was easier for programs to identify correlations that can be extrapolated to make predictions about future behaviors. It all started innocently enough, when now defunct companies like Amazon started recommending books, movies, and other products based on a consumer’s past preferences. But things got out of hand in the ’20s when this technology was used to manipulate people, such as by requiring consumers to “opt out” if they didn’t want to purchase the computer-generated list of recommended products or by assessing their previous political leanings and then automatically casting votes for them. Soon, it became impossible to “get off the grid,” and, to this day, we leave a trail of data that can be used and manipulated by corporate interests — though we still tell ourselves those uses are benign.
“The Singularity is not something to fear”
—An interview with author Roy Altman
Smarter programming had a huge impact on the workforce. At first, these programs were useful, because they predicted who was likely to leave a job, or who might be successful in a particular role. These systems morphed into IDAs (like early versions of Oscar), which were useful in anticipating problems and offering solutions. But the software had the ability to “learn” from the experiences and quickly became able to master more sophisticated tasks previously done by humans, such as writing an earnings report or analyzing market trends. Employers loved it since they could get reliable work from automation with humans playing mainly supervisory roles. But the result was that the unemployment rate skyrocketed, not only for administrative jobs. The so-called “knowledge workers,” such as doctors, lawyers, accountants, computer programmers, and stockbrokers found themselves replaced by software and joined the chronically unemployable. Eerily, the mass unemployment of the 2030s echoed that decade of the Great Depression a century earlier. However, this era came to be dubbed the Painless Depression, as the increased productivity afforded by an automated workforce created enough value in the economy for a cushy safety net. Food was easy enough to synthesize, and the U.S. government — flush with cash due to the favorable balance of trade — recognized that it had to put some money in people’s pockets in order to maintain demand for the services the automated agents produced. The bottom line was that even the unemployed and the unemployable were afforded a comfortable lifestyle thanks to the government handouts. The psychological toll of feeling unneeded, though, was devastating, and we’re still working that one out.
With so many knowledge worker tasks automated effectively, businesses naturally began placing a high value on the kinds of things software still can’t do well. With science, engineering, and math jobs mostly automated, it’s gotten to the point where if you want to work, you’d better be able to show strength in such areas as intuitive interpretation, interpersonal skills, and good old-fashioned common sense. The high-value degrees are in education and humanities and the arts.
Back in the early part of this century, IBM’s Watson system beat human champions at Jeopardy. As he went down to defeat, Ken Jennings, holder of the longest winning streak in Jeopardy history, wrote: “I for one welcome our new computer overlords.”
He got that one wrong. Fifty years later, humans still rule. Artificial intelligence is approaching human capacity, but we’re not there yet. Which brings me back to Oscar. As I mentioned earlier, he’s a sub-sentient intelligence — sentience being defined as the ability to ask why. He’s magical, amazing, and I can’t imagine living without him, but he’s no threat. Think of him as the best employee imaginable who is not interested in taking your job.
Research continues on creating the right conditions for emergent intelligent digital life. These systems would, in theory, have the ability to learn as humans do but could do so at warp speed, quickly exceeding our capacities. So, what happens when computers can truly outthink us? The issue has been raised: If a machine is as fully intelligent as a human, should it have the same rights as a human? If not, wouldn’t keeping these humanlike machines in submission be akin to slavery? Some would argue that a machine is a machine, no matter how intelligent. But others fear, as Stephen Hawking once warned, “The development of full artificial intelligence could spell the end of the human race.”
Most people today have a more benign view. I myself am coming around to thinking that not only will we have fully intelligent machines in the near future, but they are the next evolutionary step for humanity and a necessary one if the species is to survive.
You see, these new “humans” will need only electricity to live, so they’re free to leave the earth in search of a new home. If we pass the point of no return on climate change or some crazy individual or nation provokes a nuclear holocaust, finding a new home planet may be the only way for the human race to endure. Humanity, freed from the bonds of corporeality can explore the galaxy, powered by the stars. Now that we are entering an age where the very definition of life is changing, my hope and expectation is that we will embrace our new humanity.