This paper describes the development of Microsoft XiaoIce, the most popular social chatbot in the world. XiaoIce is uniquely designed as an AI companion with an emotional connection to satisfy the human need for communication, affection, and social belonging. We take into account both intelligent quotient (IQ) and emotional quotient (EQ) in system design, cast human-machine social chat as decision-making over Markov Decision Processes (MDPs), and optimize XiaoIce for long-term user engagement, measured in expected Conversation-turns Per Session (CPS). We detail the system architecture and key components including dialogue manager, core chat, skills, and an empathetic computing module. We show how XiaoIce dynamically recognizes human feelings and states, understands user intent, and responds to user needs throughout long conversations. Since the release in 2014, XiaoIce has communicated with over 660 million active users and succeeded in establishing long-term relationships with many of them. Analysis of large-scale online logs shows that XiaoIce has achieved an average CPS of 23, which is significantly higher than that of other chatbots and even human conversations.

Conversational systems have come a long way since their inception in the 1960s. After decades of research and development, we've seen progress from Eliza and Parry in the 60's and 70's, to task-completion systems as in the DARPA Communicator program in the 2000s, to intelligent personal assistants such as Siri in the 2010s, to today's social chatbots like XiaoIce. Social chatbots' appeal lies not only in their ability to respond to users' diverse requests, but also in being able to establish an emotional connection with users. The latter is done by satisfying users' need for communication, affection, as well as social belonging. To further the advancement and adoption of social chatbots, their design must focus on user engagement and take both intellectual quotient (IQ) and emotional quotient (EQ) into account. Users should want to engage with a social chatbot; as such, we define the success metric for social chatbots as conversation-turns per session (CPS). Using XiaoIce as an illustrative example, we discuss key technologies in building social chatbots from core chat to visual awareness to skills. We also show how XiaoIce can dynamically recognize emotion and engage the user throughout long conversations with appropriate interpersonal responses. As we become the first generation of humans ever living with AI, we have a responsibility to design social chatbots to be both useful and empathetic, so they will become ubiquitous and help society as a whole.

No abstract

No abstract

range (NIR-I region: 700-900 nm; NIR-II region: 1000-1700 nm). [15][16][17][18][19][20] Deep tissue penetration can be realized in this spectral range owing to the low degree level of photon absorption, scattering, and autofluorescence from tissues. [21,22] Therefore, imaging and therapeutic agents which operate in the DR to NIR region are highly desired. For CDs, strong PL in the blue and green region has been reported on numerous occasions, with PL quantum yields (QY) up to 90%, [23,24] while CDs with efficient longer wavelength emission have experienced extensive development only recently. Starting from the first report on the orange emissive CDs (PL peak at 580 nm) with a PL QY of 46%, [25] red emissive CDs with PL peak at 628 nm and PL QY of 53%, [26] and NIR emissive CDs with PL peak at 715 nm and PL QY of 43% [27] were introduced. A number of strategies, such as enlarging the π-conjugated structure, enhancing the surface oxidation, and employing proper precursors such as phenylenediamines were proved as effective for realization of such CDs. This review starts from summary of synthetic strategies toward CDs with absorption/PL at the wavelength from 650 nm and into the NIR-I and NIR-II spectral region. We then consider in detail the Stokes and anti-Stokes PL (ASPL), chemiluminescence (CL) and formation of reactive oxygen species (ROS) for the DR/NIR CDs. At the end, we provide a concise summary of already reported biomedical applications of CDs, such as in the photoacoustic (PA) imaging and photothermal therapy (PTT), photodynamic therapy (PDT), and their use as bioimaging agents and drug carriers.Carbon dots (CDs) represent a recently emerged class of luminescent materials with a great potential for biomedical theranostics, and there are a lot of efforts to shift their absorption and emission toward deep-red (DR) to near-infrared (NIR) region falling in the biological transparency window. This review offers comprehensive insights into the synthesis strategies aimed to achieve this goal, and the current approaches of modulating the optical properties of CDs over the DR to NIR region. The underlying mechanisms of their absorption, photoluminescence, and chemiluminescence, as well as the related photophysical processes of photothermal conversion and formation of reactive oxygen species are considered. The already available biomedical applications of CDs, such as in the photoacoustic imaging and photothermal therapy, photodynamic therapy, and their use as bioimaging agents and drug carriers are then shortly summarized.The ORCID identification number(s) for the author(s) of this article can be found under https://doi.org/10.1002/smll.202102325. Figure 12. a) In vivo NIR fluorescence images of a mouse stomach before (left) and after (right) gavage injection of CDs in PVP aqueous solution.Reproduced with permission. [46] Copyright 2018, Wiley-VCH. b) NIR fluorescence images of mouse tumor after intravenous injection of CDs at various time points. c) NIR fluorescence of H22 tumors that were dissected from mice at ...