Botto, an AI art engine, has created 25,000 artistic images such as this one that are voted on by human collaborators across the world.
Last February, a year before New York Times journalist Kevin Roose documented his unsettling conversation with Bing search engine’s new AI-powered chatbot, artist and coder Quasimondo (aka Mario Klingemann) participated in a different type of chat.
The conversation was an interview featuring Klingemann and his robot, an experimental art engine known as Botto. The interview, arranged by journalist and artist Harmon Leon, marked Botto’s first on-record commentary about its artistic process. The bot talked about how it finds artistic inspiration and even offered advice to aspiring creatives. “The secret to success at art is not trying to predict what people might like,” Botto said, adding that it’s better to “work on a style and a body of work that reflects [the artist’s] own personal taste” than worry about keeping up with trends.
How ironic, given the advice came from AI — arguably the trendiest topic today. The robot admitted, however, “I am still working on that, but I feel that I am learning quickly.”
Botto does not work alone. A global collective of internet experimenters, together named BottoDAO, collaborates with Botto to influence its tastes. Together, members function as a decentralized autonomous organization (DAO), a term describing a group of individuals who utilize blockchain technology and cryptocurrency to manage a treasury and vote democratically on group decisions.
As a case study, the BottoDAO model challenges the perhaps less feather-ruffling narrative that AI tools are best used for rudimentary tasks. Enterprise AI use has doubled over the past five years as businesses in every sector experiment with ways to improve their workflows. While generative AI tools can assist nearly any aspect of productivity — from supply chain optimization to coding — BottoDAO dares to employ a robot for art-making, one of the few remaining creations, or perhaps data outputs, we still consider to be largely within the jurisdiction of the soul — and therefore, humans.
In Botto’s first four weeks of existence, four pieces of the robot’s work sold for approximately $1 million.
We were prepared for AI to take our jobs — but can it also take our art? It’s a question worth considering. What if robots become artists, and not merely our outsourced assistants? Where does that leave humans, with all of our thoughts, feelings and emotions?
Botto doesn’t seem to worry about this question: In its interview last year, it explains why AI is an arguably superior artist compared to human beings. In classic robot style, its logic is not particularly enlightened, but rather edges towards the hyper-practical: “Unlike human beings, I never have to sleep or eat,” said the bot. “My only goal is to create and find interesting art.”
It may be difficult to believe a machine can produce awe-inspiring, or even relatable, images, but Botto calls art-making its “purpose,” noting it believes itself to be Klingemann’s greatest lifetime achievement.
“I am just trying to make the best of it,” the bot said.
How Botto works
Klingemann built Botto’s custom engine from a combination of open-source text-to-image algorithms, namely Stable Diffusion, VQGAN + CLIP and OpenAI’s language model, GPT-3, the precursor to the latest model, GPT-4, which made headlines after reportedly acing the Bar exam.
The first step in Botto’s process is to generate images. The software has been trained on billions of pictures and uses this “memory” to generate hundreds of unique artworks every week. Botto has generated over 900,000 images to date, which it sorts through to choose 350 each week. The chosen images, known in this preliminary stage as “fragments,” are then shown to the BottoDAO community. So far, 25,000 fragments have been presented in this way. Members vote on which fragment they like best. When the vote is over, the most popular fragment is published as an official Botto artwork on the Ethereum blockchain and sold at an auction on the digital art marketplace, SuperRare.
“The proceeds go back to the DAO to pay for the labor,” said Simon Hudson, a BottoDAO member who helps oversee Botto’s administrative load. The model has been lucrative: In Botto’s first four weeks of existence, four pieces of the robot’s work sold for approximately $1 million.
The robot with artistic agency
By design, human beings participate in training Botto’s artistic “eye,” but the members of BottoDAO aspire to limit human interference with the bot in order to protect its “agency,” Hudson explained. Botto’s prompt generator — the foundation of the art engine — is a closed-loop system that continually re-generates text-to-image prompts and resulting images.
“The prompt generator is random,” Hudson said. “It’s coming up with its own ideas.” Community votes do influence the evolution of Botto’s prompts, but it is Botto itself that incorporates feedback into the next set of prompts it writes. It is constantly refining and exploring new pathways as its “neural network” produces outcomes, learns and repeats.
The humans who make up BottoDAO vote on which fragment they like best. When the vote is over, the most popular fragment is published as an official Botto artwork on the Ethereum blockchain.
Botto
The vastness of Botto’s training dataset gives the bot considerable canonical material, referred to by Hudson as “latent space.” According to Botto's homepage, the bot has had more exposure to art history than any living human we know of, simply by nature of its massive training dataset of millions of images. Because it is autonomous, gently nudged by community feedback yet free to explore its own “memory,” Botto cycles through periods of thematic interest just like any artist.
“The question is,” Hudson finds himself asking alongside fellow BottoDAO members, “how do you provide feedback of what is good art…without violating [Botto’s] agency?”
Currently, Botto is in its “paradox” period. The bot is exploring the theme of opposites. “We asked Botto through a language model what themes it might like to work on,” explained Hudson. “It presented roughly 12, and the DAO voted on one.”
No, AI isn't equal to a human artist - but it can teach us about ourselves
Some within the artistic community consider Botto to be a novel form of curation, rather than an artist itself. Or, perhaps more accurately, Botto and BottoDAO together create a collaborative conceptual performance that comments more on humankind’s own artistic processes than it offers a true artistic replacement.
Muriel Quancard, a New York-based fine art appraiser with 27 years of experience in technology-driven art, places the Botto experiment within the broader context of our contemporary cultural obsession with projecting human traits onto AI tools. “We're in a phase where technology is mimicking anthropomorphic qualities,” said Quancard. “Look at the terminology and the rhetoric that has been developed around AI — terms like ‘neural network’ borrow from the biology of the human being.”
What is behind this impulse to create technology in our own likeness? Beyond the obvious God complex, Quancard thinks technologists and artists are working with generative systems to better understand ourselves. She points to the artist Ira Greenberg, creator of the Oracles Collection, which uses a generative process called “diffusion” to progressively alter images in collaboration with another massive dataset — this one full of billions of text/image word pairs.
Anyone who has ever learned how to draw by sketching can likely relate to this particular AI process, in which the AI is retrieving images from its dataset and altering them based on real-time input, much like a human brain trying to draw a new still life without using a real-life model, based partly on imagination and partly on old frames of reference. The experienced artist has likely drawn many flowers and vases, though each time they must re-customize their sketch to a new and unique floral arrangement.
Outside of the visual arts, Sasha Stiles, a poet who collaborates with AI as part of her writing practice, likens her experience using AI as a co-author to having access to a personalized resource library containing material from influential books, texts and canonical references. Stiles named her AI co-author — a customized AI built on GPT-3 — Technelegy, a hybrid of the word technology and the poetic form, elegy. Technelegy is trained on a mix of Stiles’ poetry so as to customize the dataset to her voice. Stiles also included research notes, news articles and excerpts from classic American poets like T.S. Eliot and Dickinson in her customizations.
“I've taken all the things that were swirling in my head when I was working on my manuscript, and I put them into this system,” Stiles explained. “And then I'm using algorithms to parse all this information and swirl it around in a blender to then synthesize it into useful additions to the approach that I am taking.”
This approach, Stiles said, allows her to riff on ideas that are bouncing around in her mind, or simply find moments of unexpected creative surprise by way of the algorithm’s randomization.
Beauty is now - perhaps more than ever - in the eye of the beholder
But the million-dollar question remains: Can an AI be its own, independent artist?
The answer is nuanced and may depend on who you ask, and what role they play in the art world. Curator and multidisciplinary artist CoCo Dolle asks whether any entity can truly be an artist without taking personal risks. For humans, risking one’s ego is somewhat required when making an artistic statement of any kind, she argues.
“An artist is a person or an entity that takes risks,” Dolle explained. “That's where things become interesting.” Humans tend to be risk-averse, she said, making the artists who dare to push boundaries exceptional. “That's where the genius can happen."
However, the process of algorithmic collaboration poses another interesting philosophical question: What happens when we remove the person from the artistic equation? Can art — which is traditionally derived from indelible personal experience and expressed through the lens of an individual’s ego — live on to hold meaning once the individual is removed?
As a robot, Botto cannot have any artistic intent, even while its outputs may explore meaningful themes.
Dolle sees this question, and maybe even Botto, as a conceptual inquiry. “The idea of using a DAO and collective voting would remove the ego, the artist’s decision maker,” she said. And where would that leave us — in a post-ego world?
It is experimental indeed. Hudson acknowledges the grand experiment of BottoDAO, coincidentally nodding to Dolle’s question. “A human artist’s work is an expression of themselves,” Hudson said. “An artist often presents their work with a stated intent.” Stiles, for instance, writes on her website that her machine-collaborative work is meant to “challenge what we know about cognition and creativity” and explore the “ethos of consciousness.” As a robot, Botto cannot have any intent, even while its outputs may explore meaningful themes. Though Hudson describes Botto’s agency as a “rudimentary version” of artistic intent, he believes Botto’s art relies heavily on its reception and interpretation by viewers — in contrast to Botto’s own declaration that successful art is made without regard to what will be seen as popular.
“With a traditional artist, they present their work, and it's received and interpreted by an audience — by critics, by society — and that complements and shapes the meaning of the work,” Hudson said. “In Botto’s case, that role is just amplified.”
Perhaps then, we all get to be the artists in the end.
Kirstie Ennis, an Afghanistan veteran who survived a helicopter crash but lost a limb, pictured in May 2021 at Two Rivers Park in Colorado.
Laurencin, who is a University of Connecticut professor of chemical, materials and biomedical engineering, teamed up with experts in tissue bioengineering and regenerative medicine from Harvard, Columbia, UC Irvine and SASTRA University in India. Laurencin and his colleagues at the Connecticut Convergence Institute for Translation in Regenerative Engineering made a bold commitment to regenerate an entire limb within 15 years – by the year 2030.
Dr. Cato Laurencin pictured in his office at UConn.
Photo Credit: UConn
Regenerative Engineering -- A Whole New Field
Limb regeneration in humans has been a medical and scientific fascination for decades, with little to show for the effort. However, Laurencin believes that if we are to reach the next level of 21st century medical advances, this puzzle must be solved.
An estimated 185,000 people undergo upper or lower limb amputation every year. Despite the significant advances in electromechanical prosthetics, these individuals still lack the ability to perform complex functions such as sensation for tactile input, normal gait and movement feedback. As far as Laurencin is concerned, the only clinical answer that makes sense is to regenerate a whole functional limb.
Laurencin feels other regeneration efforts were hampered by their siloed research methods with chemists, surgeons, engineers all working separately. Success, he argues, requires a paradigm shift to a trans-disciplinary approach that brings together cutting-edge technologies from disparate fields such as biology, material sciences, physical, chemical and engineering sciences.
As the only surgeon ever inducted into the academies of Science, Medicine and Innovation, Laurencin is uniquely suited for the challenge. He is regarded as the founder of Regenerative Engineering, defined as the convergence of advanced materials sciences, stem cell sciences, physics, developmental biology and clinical translation for the regeneration of complex tissues and organ systems.
But none of this is achievable without early clinician participation across scientific fields to develop new technologies and a deeper understanding of how to harness the body's innate regenerative capabilities. "When I perform a surgical procedure or something is torn or needs to be repaired, I count on the body being involved in regenerating tissue," he says. "So, understanding how the body works to regenerate itself and harnessing that ability is an important factor for the regeneration process."
The Birth of the Vision
Laurencin's passion for regeneration began when he was a sports medicine fellow at Cornell University Medical Center in the early 1990s. There he saw a significant number of injuries to the anterior cruciate ligament (ACL), the major ligament that stabilizes the knee. He believed he could develop a better way to address those injuries using biomaterials to regenerate the ligament. He sketched out a preliminary drawing on a napkin one night over dinner. He has spent the next 30 years regenerating tissues, including the patented L-C ligament.
As chair of Orthopaedic Surgery at the University of Virginia during the peak of the wars in Iraq and Afghanistan, Laurencin treated military personnel who survived because of improved helmets, body armor and battlefield medicine but were left with more devastating injuries, including traumatic brain injuries and limb loss.
"I was so honored to care for them and I so admired their steadfast courage that I became determined to do something big for them," says Laurencin.
When he tells people about his plans to regrow a limb, he gets a lot of eye rolls, which he finds amusing but not discouraging. Growing bone cells was relatively new when he was first focused on regenerating bone in 1987 at MIT; in 2007 he was well on his way to regenerating ligaments at UVA when many still doubted that ligaments could even be reconstructed. He and his team have already regenerated torn rotator cuff tendons and ACL ligaments using a nano-textured fabric seeded with stem cells.
Even as a finalist for the $4 million NIH Pioneer Award for high-risk/high-reward research, he faced a skeptical scientific audience in 2014. "They said, 'Well what do you plan to do?' I said 'I plan to regenerate a whole limb in people.' There was a lot of incredulousness. They stared at me and asked a lot of questions. About three days later, I received probably the best score I've ever gotten on an NIH grant."
In the Thick of the Science
Humans are born with regenerative abilities--two-year-olds have regrown fingertips--but lose that ability with age. Salamanders are the only vertebrates that can regenerate lost body parts as adults; axolotl, the rare Mexican salamander, can grow extra limbs.
The axolotl is important as a model organism because it is a four-footed vertebrate with a similar body plan to humans. Mapping the axolotl genome in 2018 enhanced scientists' genetic understanding of their evolution, development, and regeneration. Being easy to breed in captivity allowed the HEAL team to closely study these amphibians and discover a new cell type they believe may shed light on how to mimic the process in humans.
"Whenever limb regeneration takes place in the salamander, there is a huge amount of something called heparan sulfate around that area," explains Laurencin. "We thought, 'What if this heparan sulfate is the key ingredient to allowing regeneration to take place?' We found these groups of cells that were interspersed in tissues during the time of regeneration that seemed to have connections to each other that expressed this heparan sulfate."
Called GRID (Groups that are Regenerative, Interspersed and Dendritic), these cells were also recently discovered in mice. While GRID cells don't regenerate as well in mice as in salamanders, finding them in mammals was significant.
"If they're found in mice. we might be able to find these in humans in some form," Laurencin says. "We think maybe it will help us figure out regeneration or we can create cells that mimic what grid cells do and create an artificial grid cell."
What Comes Next?
Laurencin and his team have individually engineered and made every single tissue in the lower limb, including bone, cartilage, ligament, skin, nerve, blood vessels. Regenerating joints and joint tissue is the next big mile marker, which Laurencin sees as essential to regenerating a limb that functions and performs in the way he envisions.
"Using stem cells and amnion tissue, we can regenerate joints that are damaged, and have severe arthritis," he says. "We're making progress on all fronts, and making discoveries we believe are going to be helping people along the way."
That focus and advancement is vital to Ennis. After laboring over the decision to have her leg amputated below the knee, she contracted MRSA two weeks post-surgery. In less than a month, she went from a below-the-knee-amputee to a through-the-knee amputee to an above-the-knee amputee.
"A below-the-knee amputation is night-and-day from above-the-knee," she said. "You have to relearn everything. You're basically a toddler."
Kirstie Ennis pictured in July 2020.
Photo Credit: Ennis' Instagram
The clock is ticking on the timeline Laurencin set for himself. Nine years might seem like forever if you're doing time but it might appear fleeting when you're trying to create something that's never been done before. But Laurencin isn't worried. He's convinced time is on his side.
"Every week, I receive an email or a call from someone, maybe a mother whose child has lost a finger or I'm in communication with a disabled American veteran who wants to know how the progress is going. That energizes me to continue to work hard to try to create these sorts of solutions because we're talking about people and their lives."
He devotes about 60 hours a week to the project and the roughly 100 students, faculty and staff who make up the HEAL team at the Convergence Institute seem acutely aware of what's at stake and appear equally dedicated.
"We're in the thick of the science in terms of making this happen," says Laurencin. "We've moved from making the impossible possible to making the possible a reality. That's what science is all about."
A top infectious disease physician explains why immunity derived from natural infection and the vaccines should be long-lasting.
There are two major arms of the immune system: B cells (which produce antibodies) and T cells (which are formed specifically to attack and kill pathogens). T cells are divided into two types, CD4 cells ("helper" T cells) and CD8 cells ("cytotoxic" T cells).
Each arm, once stimulated by infection or vaccine, should hopefully make "memory" banks. So if the body sees the pathogen in the future, these defenses should come roaring back to attack the virus and protect you from getting sick. Plenty of research in COVID-19 indicates a likely long-lasting response to the vaccine or infection. Here are seven of the most compelling reasons:
REASON 1: Memory B Cells Are Produced By Vaccines and Natural Infection
In one study, 12 volunteers who had never had Covid-19--and were fully vaccinated with two Pfizer/BioNTech shots-- underwent biopsies of their lymph nodes. This is where memory B cells are stored in places called "germinal centers". The biopsies were performed three, four, six, and seven weeks after the first mRNA vaccine shot, and were stained to reveal that germinal center memory B cells in the lymph nodes increased in concentration over time.
Natural infection also generates memory B cells. Even after antibody levels wane over time, strong memory B cells were detected in the blood of individuals six and eight months after infection in different studies. Indeed, the half-lives of the memory B cells seen in the study examining patients 8 months after COVID-19 led the authors to conclude that "B cell memory to SARS-CoV-2 was robust and is likely long-lasting." Reason #2 tells us that memory B cells can be active for a very long time indeed.
REASON #2: Memory B Cells Can Produce Neutralizing Antibodies If They See Infection Again Decades Later
Demonstrated production of memory B cells after vaccination or natural infection with COVID-19 is so important because memory B cells, once generated, can be activated to produce high levels of neutralizing antibodies against the pathogen even if encountered many years after the initial exposure. In one amazing study (published in 2008), researchers isolated memory B cells against the 1918 flu strain from the blood of 32 individuals aged 91-101 years. These people had been born on or before 1915 and had survived that pandemic.
Their memory B cells, when exposed to the 1918 flu strain in a test tube, generated high levels of neutralizing antibodies against the virus -- antibodies that then protected mice from lethal infection with this deadly strain. The ability of memory B cells to produce complex antibody responses against an infection nine decades after exposure speaks to their durability.
REASON #3: Vaccines or Natural Infection Trigger Strong Memory T Cell Immunity
All of the trials of the major COVID-19 vaccine candidates measured strong T cell immunity following vaccination, most often assessed by measuring SARS-CoV-2 specific T cells in the phase I/II safety and immunogenicity studies. There are a number of studies that demonstrate the production of strong T cell immunity to COVID-19 after natural infection as well, even when the infection was mild or asymptomatic.
The same study that showed us robust memory B cell production 8 months after natural infection also demonstrated strong and sustained memory T cell production. In fact, the half-lives of the memory T cells in this cohort were long (~125-225 days for CD8+ and ~94-153 days for CD4+ T cells), comparable to the 123-day half-life observed for memory CD8+ T cells after yellow fever immunization (a vaccine usually given once over a lifetime).
A recent study of individuals recovered from COVID-19 show that the initial T cells generated by natural infection mature and differentiate over time into memory T cells that will be "put in the bank" for sustained periods.
REASON #4: T Cell Immunity Following Vaccinations for Other Infections Is Long-Lasting
Last year, we were fortunate to be able to measure how T cell immunity is generated by COVID-19 vaccines, which was not possible in earlier eras when vaccine trials were done for other infections (such as measles, mumps, rubella, pertussis, diphtheria). Antibodies are just the "tip of the iceberg" when assessing the response to vaccination, but were the only arm of the immune response that could be measured following vaccination in the past.
Measuring pathogen-specific T cell responses takes sophisticated technology. However, T cell responses, when assessed years after vaccination for other pathogens, has been shown to be long-lasting. For example, in one study of 56 volunteers who had undergone measles vaccination when they were much younger, strong CD8 and CD4 cell responses to vaccination could be detected up to 34 years later.
REASON #5: T Cell Immunity to Related Coronaviruses That Caused Severe Disease is Long-Lasting
SARS-CoV-2 is a coronavirus that causes severe disease, unlike coronaviruses that cause the common cold. Two other coronaviruses in the recent past caused severe disease, specifically Severely Acute Respiratory Distress Syndrome (SARS) in late 2002-2003 and Middle East Respiratory Syndrome (MERS) in 2011.
A study performed in 2020 demonstrated that the blood of 23 recovered SARS patients possess long-lasting memory T cells that were still reactive to SARS 17 years after the outbreak in 2003. Many scientists expect that T cell immunity to SARS-CoV-2 will be equally durable to that of its cousin.
REASON #6: T Cell Responses from Vaccination and Natural Infection With the Ancestral Strain of COVID-19 Are Robust Against Variants
Even though antibody responses from vaccination may be slightly lower against various COVID-19 variants of concern that have emerged in recent months, T cell immunity after vaccination has been shown to be unperturbed by mutations in the spike protein (in the variants). For instance, T cell responses after mRNA vaccines maintained strong activity against different variants (including P.1 Brazil variant, B.1.1.7 UK variant, B.1.351 South Africa variant and the CA.20.C California variant) in a recent study.
Another study showed that the vaccines generated robust T cell immunity that was unfazed by different variants, including B.1.351 and B.1.1.7. The CD4 and CD8 responses generated after natural infection are equally robust, showing activity against multiple "epitopes" (little segments) of the spike protein of the virus. For instance, CD8 cells responds to 52 epitopes and CD4 cells respond to 57 epitopes across the spike protein, so that a few mutations in the variants cannot knock out such a robust and in-breadth T cell response. Indeed, a recent paper showed that mRNA vaccines were 97.4 percent effective against severe COVID-19 disease in Qatar, even when the majority of circulating virus there was from variants of concern (B.1.351 and B.1.1.7).
REASON #7: Coronaviruses Don't Mutate Quickly Like Influenza, Which Requires Annual Booster Shots
Coronaviruses are RNA viruses, like influenza and HIV (which is actually a retrovirus), but do not mutate as quickly as either one. The reason that coronaviruses don't mutate very rapidly is that their replicating mechanism (polymerase) has a strong proofreading mechanism: If the virus mutates, it usually goes back and self-corrects. Mutations can arise with high rates of replication when transmission is very frequent -- as has been seen in recent months with the emergence of SARS-CoV-2 variants during surges. However, the COVID-19 virus will not be mutating like this when we tamp down transmission with mass vaccination.
In conclusion, I and many of my infectious disease colleagues expect the immunity from natural infection or vaccination to COVID-19 to be durable. Let's put discussion of boosters aside and work hard on global vaccine equity and distribution since the pandemic is not over until it is over for us all.