This scenario isn’t that far from reality, as AI chatbots are already taking over online commerce. By 2025, 95% of companies will have an AI chatbot, according to Finance Digest. AI might not be sentient yet, but it can be programmed to express emotions.

Humans displaying positive emotions in customer service interactions have long been known to improve customer experience, but researchers at the Georgia Institute of Technology’s Scheller College of Business wanted to see if this also applied to AI. They conducted experimental studies to determine if positive emotional displays improved customer service and found that emotive AI is only appreciated if the customer expects it, and it may not be the best avenue for companies to invest in.

“It is commonly believed and repeatedly shown that human employees can express positive emotion to improve customers’ service evaluations,” said Han Zhang, the Steven A. Denning Professor in Technology & Management. “Our findings suggest that the likelihood of AI’s expression of positive emotion to benefit or hurt service evaluations depends on the type of relationship that customers expect from the service agent.”

The researchers presented their findings in the paper, “Bots With Feelings: Should AI Agents Express Positive Emotion in Customer Service?,” in Information Systems Research in December.

Studying AI Emotion

The researchers conducted three studies to expand the understanding of emotional AI in customer service transactions. Although they changed the participants and scenario in each study, AI chatbots imbued with emotion used positive emotional adjectives, such as excited, delighted, happy, or glad. They also deployed more exclamation points.

The first study focused on whether customers responded more favorably to positive emotion if they knew the customer agent was a bot or person. Participants were told they were seeking help for a missing item in a retail order. The 155 participants were then randomly assigned to four different scenarios: human agents with neutral emotion, human agents with positive emotion, bots with neutral emotion, and bots with positive emotion. Then they asked participants about service quality and overall satisfaction. The results indicated that positive emotion was more beneficial when human agents exhibited it, but it had no effect when bots exhibited it.

The second study examined if customers’ personal expectations determined their reaction to the bot. In this scenario, the 88 participants imagined returning a textbook and were randomly assigned to either emotion-positive or emotion-neutral bots. After chatting with the bot, they were asked to rate if they were communal (social) oriented or exchange (transaction) oriented on a scale. If the participant was communal-focused, they were more likely to appreciate the positive emotional bot, but if they expected the exchange as merely transactional, the emotionally positive bot made their experience worse.

“Our work enables businesses to understand the expectations of customers exposed to AI-provided services before they haphazardly equip AIs with emotion-expressing capabilities,” Zhang said.

The final study explored why a bot’s positive emotion influences customer emotions, following 177 undergraduate students randomly assigned to emotive or non-emotive bots. The results explained why positive bots have less of an effect than anticipated. Because customers do not expect machines to have emotions, they can react negatively to emotion in a bot.

The results across the studies show that using positive emotion in chatbots is challenging because businesses don’t know a customer’s biases and expectations going into the interaction. A happy chatbot could lead to an unhappy customer.

“Our findings suggest that the positive effect of expressing positive emotion on service evaluations may not materialize when the source of the emotion is not a human,” Zhang said. “Practitioners should be cautious about the promises of equipping AI agents with emotion-expressing capabilities.”

The children have Artemis-SCID, a very rare genetic disorder that is typically treated with a bone marrow transplant from a healthy donor, ideally a matched brother or sister. The new gene therapy allows researchers to treat newly diagnosed babies with their own cells — adding a healthy copy of the Artemis gene to the baby’s harvested marrow stem cells, then infusing the corrected stem cells back into their bodies — in hopes of avoiding many of the short- and long-term complications of the standard treatment, including death.

The children in the trial — all under the age of 5 — are living at home with their families, attending daycare and preschool, playing outside, and living normal lives, said Mort Cowan, MD, UCSF pediatrics professor and the trial’s lead investigator.

“Already, the course of their illness is so much better than with the typical treatment,” said Cowan, who has treated more than 30 children with Artemis-SCID using standard bone marrow transplants. “I’ve never seen results like this in any of the other kids. It’s amazing.”

Gene correction has been used before in patients with other genetic forms of SCID, but its use in Artemis-SCID is significant because these patients usually respond more poorly to standard bone marrow transplants. Complications can include rejecting the marrow graft, graft-vs.-host disease — in which the donor T cells attack the recipient’s tissues — chronic infections leading to organ damage, stunted growth, and premature death.

Signs of Stronger Immunity

The first outcome of the Phase I/II trial involved the safe transfusion of gene-corrected cells that would differentiate into white blood cells by 42 days after infusion. Researchers theorized patients would need less chemotherapy to prepare their marrow for transfusion when their own cells were being used; thus only 25% of a full dose of busulfan was administered. The second outcome was T-cell reconstitution at 12 months, a measurement of the strength of the immune system.

All 10 patients were safely transfused with their own gene-corrected stem cells that gave rise to corrected peripheral blood cells within 42 days. All 10 were growing their own T cells and B cells by 12 weeks, and four of nine (excluding a patient who received a second treatment) achieved full T-cell immune reconstitution by 12 months. Four of nine also achieved full B cell immunity by 24 months, allowing them to discontinue immunoglobulin replacement and receive standard childhood vaccinations. An additional three patients, who were followed for fewer than 24 months, had promising B cell development when compared to previous outcomes for donor-transplanted patients.

One child required a second infusion of gene-corrected bone marrow due to a persistent infection with cytomegalovirus prior to gene therapy but is now infection free with good T- and B-cell immunity. “All of the results are better than those previously seen with Artemis-SCID patients who received donor bone marrow transplants,” noted Jennifer Puck, MD, UCSF pediatrics professor and co-lead investigator in the study.

“Having patients in the trial achieve full T-cell immunity is outstanding. B-cell recovery takes longer, but so far it looks as if the patients also have a far better chance for B-cell reconstitution than they would with a regular bone marrow transplant,” Puck said. “Successfully using less chemotherapy is also a big win, minimizing the harmful side effects of full dose busulfan in small infants.”

Better B-cell immunity could help avoid issues such as chronic lung disease that often develop later in childhood for Artemis-SCID patients who receive a standard bone marrow transplant, Cowan added.

The children in the trial are currently between the ages of 18 months and 4.5 years; nine were born in the U.S. and were diagnosed following newborn screening for SCID; one was born in Canada and diagnosed at five months of age with clinical illness. Four patients are of Navajo/Apache Native American descent, where the Artemis-SCID mutation is more common. Median follow-up was 31.2 months. At the time of study publication, six patients had been followed for at least 24 months.

“We’re pioneering gene therapy in this very rare disease right now, but we are using techniques that can be exported to other situations and can help many other conditions worldwide,” said Puck. “Every new innovation happens one patient at a time.”

Co-authors: Also contributing to this research from UCSF were: Jason Yu, PhD; Carol Fraser-Browne, BA; Ukina Sanford, MS; Misako Kawahara, BA; Wendy Chan, BS; Shivali Chag, MS; and Robert Currier, PhD, of the UCSF Department of Pediatrics; Jess Oh, MS, of the UCSF Benioff Children’s Hospital San Francisco; Jasmeen Dara, MD; Janelle Facchino, NP; Christopher Dvorak, MD, of the UCSF Department of Pediatrics and UCSF Benioff Children’s Hospital San Francisco; Joan Hilton, DSc, MPH, of the UCSF Department of Epidemiology and Biostatistics; and Janel Long-Boyle, PharmD, PhD, of UCSF’s Department of Pediatrics, School of Pharmacy and UCSF Benioff Children’s Hospital San Francisco. Please refer to the paper for additional co-authors.

While most people with MS are first diagnosed with relapsing-remitting MS, marked by symptom flare-ups followed by periods of remission, many people with relapsing-remitting MS eventually transition to secondary progressive MS, which does not have wide swings in symptoms but instead a slow, steady worsening of the disease.

“Hematopoietic stem cell transplants have been previously found to delay disability in people with relapsing-remitting MS, but less is known about whether such transplants could help delay disability during the more advanced stage of the disease,” said study author Matilde Inglese, MD, PhD, of the University of Genoa in Italy and a member of the American Academy of Neurology. “Our results are encouraging, because while current treatments for secondary progressive MS have modest or small benefits, our study found stem cell transplants may not only delay disability longer than many other MS medications, they may also provide a slight improvement in symptoms.”

The retrospective study included 79 people with active secondary progressive MS who received stem cell transplants and 1,975 people from the Italian MS registry who were treated with MS drugs. All received treatment after being diagnosed with active secondary progressive MS. The two groups were matched for age, sex and level of disability. Drugs included beta-interferons, azathioprine, glatiramer acetate, mitoxantrone, fingolimod, natalizumab, methotrexate, teriflunomide, cyclophosphamide, dimethyl fumarate and alemtuzumab.

Participants’ level of disability was measured on the Expanded Disability Status Scale, a common method to quantify disability with scores ranging from 0, no symptoms, to 10 points, death due to MS. Participants were assessed at various time points over 10 years.

At the beginning of the study, participants had a median score of 6.5 for both those who received transplants and those receiving the medications. Scores of 6.0 are defined as needing to use a cane or brace intermittently or on one side to walk about 100 meters with or without resting. Scores of 6.5 are defined as needing to use a cane or brace constantly on both sides to walk about 20 meters without resting.

Five years into the study, researchers found 62% of the people who had stem cell transplants experienced no worsening of their MS disability compared to 46% of those who took medications.

Also, at five years, researchers found people who received stem cell transplants were more likely to see sustained improvements over time, with 19% experiencing less disability than at the start of the study, compared to just 4% of people taking medications.

Over 10 years, the disability score for people who had stem cell transplants decreased by an average of 0.01 points per year, signifying less disability, while the average score for people taking medications increased by 0.16 points per year, an increase in disability.

“Our study shows that hematopoietic stem cell transplants were associated with a slowing of disability progression and a higher likelihood of disability improvement compared to other therapies,” said Inglese. “While these results are encouraging, they are not applicable to patients with secondary progressive MS who do not have signs of inflammatory disease activity; more research is needed in larger groups of people to confirm our findings.”

A limitation of the study is that it was retrospective and observational, and does not prove cause and effect. It only suggests an association. The study also did not include people taking the MS drugs siponimod, cladribine, ocrelizumab, ofatumumab, or rituximab.

The study was funded by the Italian Multiple Sclerosis Foundation.