Photo from MIT Technology Review.
Companies are turning to artificial intelligence to automate tasks, from designing clothes to testing new medications.
AI can analyze a person's data, including what they look like, what they buy, and their health information. Based on this, AI can predict things like how clothes would look on them, how they might answer a question, or even how a disease might affect them. This AI-generated information, sometimes called a digital twin, is being used for many different purposes already.
AI is making waves in several industries. In fashion, Los Angeles startup AI Fashion uses real models' photos to create entirely new AI-generated images for them, showcasing various clothing options for online stores and campaigns. Brox AI, another startup, has built a virtual world of 27,000 individuals with detailed profiles including shopping habits and brand preferences. This allows companies to conduct virtual focus groups with these AI profiles, asking questions about products or marketing strategies to gauge potential customer responses. Healthcare is also seeing an AI revolution. San Francisco's Unlearn leverages AI to create digital twins of people based on their health data. These digital twins can predict how a disease might progress in that specific person, allowing researchers to design more efficient and effective clinical trials.
There's a flip side to the potential job displacement caused by AI. These companies creating digital twins emphasize the continued importance of human involvement. People can be compensated for contributing their data to create these AI counterparts. For businesses, digital people offer a way to expand their reach and reduce costs. It's a win-win situation where humans are valued for their data while companies gain efficiency through AI. However, the long-term impact on the human workforce remains a question.
The future might be filled with digital versions of ourselves. Consulting giant Gartner calls this technology "digital humans" and predicts companies might create one for every customer within a decade. However, it's not all sunshine and roses. Consumers might revolt if companies don't handle the data collection and usage ethically. Despite these concerns, businesses are already dipping their toes into this territory, looking to leverage AI to create digital representations of us, potentially for marketing or even monetization.
In fashion
Women's clothing brand Anne Klein is embracing AI for their fashion shoots. They're testing technology from AI Fashion that can create entire photoshoots based on just pictures of real models. This AI technology is a combination of unique, in-house developments and cutting-edge open-source models available to the public.
There's a growing demand from customers for a more personalized shopping experience, but they also want to see how clothes look in different settings. Doug Weiss, a senior executive at WHP Global (parent company of Anne Klein), sees AI as the key to achieving both. AI allows them to personalize recommendations at scale, while also generating images of the clothing in various environments, giving customers a more well-rounded idea of the product.
Doug Weiss, a higher-up at the company that owns Anne Klein, assures us that AI won't eliminate photoshoots altogether. He sees AI Fashion's tech as more of a collaborator. It allows them to generate the wide variety of content that shoppers expect nowadays, like seeing clothes in different environments. This way, Anne Klein can cater to the two things customers want: personalized recommendations and the ability to see how clothes would look in various situations.
The world of fashion is embracing AI for content creation. Several startups offer services that use AI to generate images of clothing on models, with some companies even creating entirely AI-generated models. This approach has its drawbacks though, with critics arguing that it could replace real models and take away jobs in the fashion industry.
AI Fashion stands out from the crowd with its unique approach. Founded in 2020 by Daniel Citron (a former creative lead at Google) and John Chirikjian (the company's CTO), AI Fashion places real human models at the heart of their process. This sets them apart from competitors who might solely rely on AI-generated models.
AI Fashion acknowledges that compensation for models depends on various factors like the brand involved, the number of images needed, and the model's popularity. Importantly, they also highlight that models have the freedom to reject campaigns they don't feel comfortable with.
While Doug Weiss anticipates AI Fashion's tool to lead to more personalized shopping experiences and cost savings, he admits it's still early to determine the exact financial benefits. This suggests they're using AI cautiously, optimistic about its potential but aware that its true impact on their bottom line needs more time to assess.
Brox AI offers a game-changer for market research. Their focus group tool allows companies to gather valuable consumer insights without the hefty costs and lengthy setup times associated with traditional focus groups. This innovative approach streamlines the process, potentially saving businesses significant time and money.
The co-founder and CEO of Brox, Hamish Brocklebank, stated that their tool utilizes digital replicas of 27,000 actual individuals.
Brocklebank explained that they have detailed knowledge of individuals' shopping habits, purchase preferences, and interests, obtained primarily through extensive interviews.
Using interview data, Brox's exclusive AI algorithm can provide insights into questions such as whether a woman in her thirties would be willing to pay a 10% higher fee for a streaming service subscription. Participants received compensation ranging from $20 to $150, depending on the number of interviews they took part in, according to Brocklebank.
With the tool, companies can input inquiries about which offers might appeal to specific consumers, their sensitivity to price changes, or the factors that might encourage them to subscribe to new services.
Brocklebank mentioned that the tool's annual cost ranges from $25,000 to several hundred thousand dollars, depending on companies' usage. He highlighted that this presents potential savings for companies that currently spend millions annually on focus groups.
During Clinical Trials
Far away from the fashion photoshoots or corporate focus groups, startup Unlearn is using AI to generate digital twins of people that predict how a particular disease might progress over time.
The company, founded in 2017, has raised over $130 million and has 69 employees.
Usually, during clinical trials, one group of participants is administered an experimental drug and closely monitored to evaluate its side effects and efficacy. Meanwhile, a second group is given a placebo and monitored to observe the natural progression of the disease without the experimental drug.
Charles Fisher, the CEO, explained that Unlearn collects initial health data from participants, processes it through a custom model trained on extensive clinical data, and creates a digital replica for each individual. This replica predicts how the disease would progress if the individual were in the placebo group.
He mentioned that employing a digital twin for the placebo group allows more actual individuals to be in the experimental group, thus granting them access to the potentially life-saving treatment.
Fisher noted that a primary reason patients decline to participate in trials is their reluctance to be randomly assigned to placebo groups. By using digital twins for placebos, everyone in the trial gains access to the experimental therapy, fulfilling the primary motivation for their participation.
Fisher mentioned that patients would agree to the creation of digital twins as part of their standard consent process for participating in the clinical trial, and they probably wouldn't receive any extra compensation for it.
Kasper Roet, co-founder and CEO of biotechnology company QurAlis, suggested that this technology could be particularly significant for diseases like ALS, also known as Lou Gehrig’s disease, where patients usually succumb within three years.
QurAlis is exploring and developing precision medicine therapies for ALS, frontotemporal dementia, and other neurodegenerative conditions. Roet mentioned that the company plans to initiate testing of Unlearn's technology, alongside traditional human placebo groups, as early as next year.
Roet expressed regret that currently, they have to administer a placebo drug to patients with a life-threatening illness just to conduct the trial. He added that there is still more work to be done before technology from companies like Unlearn can completely remove the need for placebo groups.
While we aren't there yet, that's the ultimate goal, according to Roet. He expressed optimism that they will achieve it eventually.