ELIZA, created by Joseph Weizenbaum at MIT in the early 1960s, is considered the earliest chatbot. He programmed it in Michigan Algorithm Decoder-Symmetric List Processor (MAD-SLIP) on MIT’s Compatible Time-Sharing System (CTSS) operating system, on an IBM 7094. We discovered an original ELIZA printout in Prof. Weizenbaum’s papers at MIT’s Institute Archives, including an early version of its famous DOCTOR script, a nearly complete version of the MAD-SLIP code, and various support functions in MAD and Fortran Assembly Program. Here we describe the reconstruction and reanimation of this original ELIZA on a restored CTSS, running on an emulated IBM 7094. The entire stack is open source, so that any user of a Unix-like operating system can run the world’s earliest chatbot on their own version of a pioneering time-sharing system.

ELIZA, created by Joseph Weizenbaum at MIT in the early 1960s, is usually considered the world's first chatbot. It was developed in MAD-SLIP on MIT's CTSS, the world's first time-sharing system, on an IBM 7094. We discovered an original ELIZA printout in Prof. Weizenbaum's archives at MIT, including an early version of the famous DOCTOR script, a nearly complete version of the MAD-SLIP code, and various support functions in MAD and FAP. Here we describe the reanimation of this original ELIZA on a restored CTSS, itself running on an emulated IBM 7094. The entire stack is open source, so that any user of a unix-like OS can run the world's first chatbot on the world's first time-sharing system.

The Linac Coherent Light Source (LCLS) is the world's first x-ray free electron laser. It is a scientific user facility operated by the SLAC National Accelerator Laboratory, at Stanford, for the U.S. Department of Energy. As beam time at LCLS is extremely valuable and limited, experimental efficiency-getting the most high quality data in the least time-is critical. Our overall project employs cognitive engineering methodologies with the goal of improving experimental efficiency and increasing scientific productivity at LCLS by refining experimental interfaces and workflows, simplifying tasks, reducing errors, and improving operator safety and stress. Here, we describe a multi-agent, multi-scale computational cognitive interaction model of instrument operations at LCLS. Our model simulates the aspects of human cognition at multiple cognitive and temporal scales, ranging from seconds to hours, and among agents playing multiple roles, including instrument operator, real time data analyst, and experiment manager. The model can roughly predict impacts stemming from proposed changes to operational interfaces and workflows. Example results demonstrate the model's potential in guiding modifications to improve operational efficiency. We discuss the implications of our effort for cognitive engineering in complex experimental settings and outline future directions for research. The model is open source, and the videos of the supplementary material provide extensive detail.

ELIZA, often considered the world's first chatbot, was written by Joseph Weizenbaum in the early 1960s. Weizenbaum did not intend to invent the chatbot, but rather to build a platform for research into human-machine conversation and the important cognitive processes of interpretation and misinterpretation. His purpose was obscured by ELIZA's fame, resulting in large part from the fortuitous timing of it's creation, and it's escape into the wild. In this paper I provide a rich historical context for ELIZA's creation, demonstrating that ELIZA arose from the intersection of some of the central threads in the technical history of AI. I also briefly discuss how ELIZA escaped into the world, and how its accidental escape, along with several coincidental turns of the programming language screws, led both to the misapprehension that ELIZA was intended as a chatbot, and to the loss of the original ELIZA to history for over 50 years.

Purpose:xDECIDE is a clinical decision support system, accessed through a web portal and powered by a "Human-AI Team," that offers oncology health care providers with treatment options personalized for their cancer patients and outcomes tracking through an observational research protocol. This article describes the xDECIDE process and the artificial intelligence (AI)-assisted technologies that ingest electronic medical records, generate a structured personal health record, standardize clinico-genomic features, and produce ranked treatment options based on clinical evidence, expert insights, and real-world outcomes generated by the system itself. Methods:Patients enroll directly into the IRB-approved pan-cancer XCELSIOR registry (NCT03793088). Patient consent permits data aggregation, continuous learning from clinical outcomes, and sharing of limited datasets for research. xDECIDE aggregates and processes medical records with natural language processing and machine learning to generate a structured care summary with a standardized list of patient features. These features are utilized by an ensemble of AI-based models called xCORE (xCures Option Ranking Engine) to create a ranked list of treatment options. The output of xCORE is reviewed by molecular pharmacologists and oncologists in a virtual tumor board (VTB) setting to create a report of treatment options and supporting rationales, individualized to the patient. Treating physicians use an interactive portal to view these data and reports and to continuously monitor their patients' information. Results:At the time of writing, over 7000 patients have enrolled in XCELSIOR, including over 1000 with central nervous system cancers, over 800 with pancreatic cancer, and over 300 patients each with lung, breast, and colorectal cancers. Over 350 VTBs have been performed for patients across indications including glioma, pancreatic cancer, and fibrolamellar carcinoma, with >450 therapeutic options discussed and over 2000 consensus rationales delivered. Over 500 treatment rationale statements ("rules") have been encoded to reference discrete patient features to improve algorithm decision-making. Conclusion:xDECIDE clinical decision support can aid oncologists in their practice of medicine. The system identifies potentially effective treatment options individualized for each patient from the integration of real-world evidence, human expert knowledge and opinion, and scientific and clinical publications and databases, and offers a platform to learn from the experience of every patient.

Randomized controlled trials (RCTs) offer a clear causal interpretation of treatment effects, but are inefficient in terms of information gain per patient. Moreover, because they are intended to test cohort-level effects, RCTs rarely provide information to support precision medicine, which strives to choose the best treatment for an individual patient. If causal information could be efficiently extracted from widely available real-world data, the rapidity of treatment validation could be increased, and its costs reduced. Moreover, inferences could be made across larger, more diverse patient populations. We created a "virtual trial" by fitting a multilevel Bayesian survival model to treatment and outcome records self-reported by 451 brain cancer patients. The model recovers group-level treatment effects comparable to RCTs representing over 3200 patients. The model additionally discovers the feature-treatment interactions needed to make individual-level predictions for precision medicine. By learning from heterogeneous real-world data, virtual trials can generate more causal estimates with fewer patients than RCTs, and they can do so without artificially limiting the patient population. This demonstrates the value of virtual trials as a complement to large randomized controlled trials, especially in highly heterogeneous or rare diseases.

The problem of interpreting or aggregating multiple rankings is common to many real-world applications. Perhaps the simplest and most common approach is a weighted rank aggregation, wherein a (convex) weight is applied to each input ranking and then ordered. This paper describes a new tool for visualizing and displaying ranking information for the weighted rank aggregation method. Traditionally, the aim of rank aggregation is to summarize the information from the input rankings and provide one final ranking that hopefully represents a more accurate or truthful result than any one input ranking. While such an aggregated ranking is, and clearly has been, useful to many applications, it also obscures information. In this paper, we show the wealth of information that is available for the weighted rank aggregation problem due to its structure. We apply weight set decomposition to the set of convex multipliers, study the properties useful for understanding this decomposition, and visualize the indifference regions. This methodology reveals information--that is otherwise collapsed by the aggregated ranking--into a useful, interpretable, and intuitive decision support tool. Included are multiple illustrative examples, along with heuristic and exact algorithms for computing the weight set decomposition.

Abstract“Hypernormal science” has minimal potential for contestation on matters of principle and practice so that information exchange can be unproblematic. Sciences comprise hypernormal domains and more contestable “normal” domains where knowledge diffusion, like acquiring linguistic fluency, depends on face-to-face interaction. Hypernormal domains belonging to molecular biology are contrasted with normal domains in gravitational wave detection physics. Sciences as a whole should not be confused with their typical domains. The analysis has immediate implications for proposed transitions out of the Covid-19 lockdown, proposed solutions to the replication crisis, and, perhaps, our understanding of the early development of social studies of science.

xCures operates a direct-to-patient, real-world evidence platform for decentralized clinical research. The platform leverages a nationwide observational research protocol (XCELSIOR, NCT03793088) to aggregate, normalize, and analyze N-of-1 clinical outcomes to continuously learn from and inform treatment decisions. Individual data elements are extracted directly from medical documents such as clinic notes, and radiology, genomics, and pathology reports. The data elements are standardized to established biomedical ontologies and stored in a validated and part 11 compliant electronic database, suitable for statistical analyses and regulatory filings. This permits comparison of patient outcomes across institutions and removes the burden of data entry from oncologists and their staff. As an extremely efficient real-world data solution, we have utilized this platform to accelerate both academic- and commercial-sponsored clinical research, prospectively integrating diagnostics and algorithms with interventional treatments. For each patient that participates in XCELSIOR, artificial intelligence-powered clinical decision support algorithms suggest testing and treatment options. These options and supporting treatment rationales are sourced from key opinion leaders, tumor boards, clinical researchers, practicing oncologists, and published literature, and ranked using the real-world outcomes data from the registry. At the conference, we will present an overview of this real-time learning infrastructure and report on clinical case studies for pharma and non-profit groups including over 75 virtual tumor boards and real-world evidence generated from over 150 patients with CNS cancers that we have helped in partnership with Cancer Commons and The Musella Foundation for Brain Tumor Research and Education. Outcomes analyses stratified by therapeutic interventions and biomarkers will be reported, including frequency of adverse events, time to treatment failure, time to disease progression, and overall survival. Interventions include standard-of-care chemotherapies as well as therapies accessed by clinical trial, expanded access, and off-label prescription.

e16735 Background: We use a real-world data approach to report on safety and benefits on metastatic pancreatic cancer pts who were treated with a MEK inhibitor plus hydroxychloroquine (HCQ) after exhausting all other treatment options. MEK inhibition acts on the KRAS pathway, which in turn increases autophagy as a resistance mechanism, furthermore, HCQ inhibits autophagy causing a cytotoxic effect. This combination was shown to diminish tumor volume in xenograft mouse models and a partial response in one heavily pre-treated patients was reported. Methods: XCELSIOR is an IRB approved, patient-centric, real-world data and outcomes registry for developing operational and analytic methods in precision oncology. Searching the XCELSIOR database, we identified 14 pts for whom this regimen had been considered. As part of their participation in XCELSIOR, these patients shared access to their full medical records, which were collected, processed, and abstracted into a 21 CFR 11 compliant database for analysis. We additionally collected de-identified data on 12 pts treated with this combination from five academic centers. Three more patients are expected to start treatment soon. Results: Between March 2018 and January 2020, 15 patients treated with the trametinib/HCQ combination and 3 patients treated with cobimetinib/HCQ were identified in XCELSIOR and five academic institutions. The median age at diagnosis was 64 (range 43-74) and 56% were male. For patients treated with trametinib/HCQ, the median time on treatment was 67 days (range 5-172 days), 11 patients were treated for more than 30 days (median time 97 days). The median PFS for this group was 2.9 months and the median OS was 7.4 months. The clinical benefit rate was 60% for the 10 evaluable patients treated with trametinib/HCQ, 1 patient had a partial response (previously published), 5 had stable disease (for at least 8 weeks) and 4 had progressive disease (physician reported). 2/3 patients treated with cobimetinib/HCQ were on treatment for more than 30 days and all three had progressive disease within 7 weeks. The most common side effects were Grade 1 fatigue and Grade 1/2 rash for both combinations. An additional 3 patients will start treatment soon and will be included in the analysis. Conclusions: Combinatorial MEK and autophagy inhibition was well tolerated in heavily treated metastatic pancreatic cancer patients. Trametinib/HCQ demonstrates some clinical benefit for this group. We demonstrate the feasibility of utilizing real-world data in precision oncology. Clinical trial information: NCT03793088 .

Global Cumulative Treatment Analysis (GCTA) is a novel clinical research model combining expert knowledge, and treatment coordination based upon global information-gain, to treat every patient optimally while efficiently searching the vast space that is the realm of cancer research.

Background We describe a prototype implementation of a platform that could underlie a Precision Oncology Rapid Learning system. Results We describe the prototype platform, and examine some important issues and details. In the Appendix we provide a complete walk-through of the prototype platform. Conclusions The design choices made in this implementation rest upon ten constitutive hypotheses, which, taken together, define a particular view of how a rapid learning medical platform might be defined, organized, and implemented.

In this paper, we analyse past approaches in designing and marketing Augmented Reality apps or devices, and we discover key issues that prevented their mainstream adoption. We offer insights into novel approaches to marketing consumer technology products; in particular Augmented Reality devices. We analyse marketing strategies in order to understand which have been successful and which problems should be avoided in future campaigns. Although other papers have discussed some of these topics, none of them has discussed them with an interdisciplinary lens. We also attempt to understand how recent Augmented Reality and Virtual Reality devices have been perceived in society and reported by media. This research includes human-computer interaction expert interviews, historical evolution models for consumer technology, and socio-cultural and political analyses.

In their Policy Forum “Countering imprecision in precision medicine” (29 July, p. [448][1]), S. P. Hey and A. S. Kesselheim discuss the many combinations of biomarkers and treatments that drive precision medicine research. They propose that “Funding agencies could award responsibility for

In addition to visiting high profile sites such as Facebook and Google, web users often visit more modest sites, such as those operated by bloggers, or by local organizations such as schools. Such sites, which we call "Just Plain Sites" (JPSs) are likely to inadvertently represent greater privacy risks than high profile sites by virtue of being unable to afford privacy expertise. To assess the prevalence of the privacy risks to which JPSs may inadvertently be exposing their visitors, we analyzed a number of easily observed privacy practices of such sites. We found that many JPSs collect a great deal of information from their visitors, share a great deal of information about their visitors with third parties, permit a great deal of tracking of their visitors, and use deprecated or unsafe security practices. Our goal in this work is not to scold JPS operators, but to raise awareness of these facts among both JPS operators and visitors, possibly encouraging the operators of such sites to take greater care in their implementations, and visitors to take greater care in how, when, and what they share.

Often communicate private data in informal settings such as email, where we trust that the recipient shares our assumptions regarding the disposition of this data. Sometimes we informally express our desires in this regard, but there is no formal means in such settings to make our wishes explicit, nor to hold the recipient accountable. Here we describe a system and prototype implementation called Recipient-Accountable Private Personal Data, which lets the originator express his or her privacy desires regarding data transmitted in email, and provides some accountability. Our method only assumes that the recipient is reading the email online, and on an email reader that will execute HTML and JavaScript.

Large scientific knowledge bases (KBs) are bound to contain inconsistencies and under-specified knowledge. Inconsistencies are inherent because the approach to modeling certain phenomena evolves over time, and at any given time, contradictory approaches to modeling a piece of domain knowledge may simultaneously exist in the KB. Underspecification is inherent because a large, complex KB is rarely fully specified, especially when authored by domain experts who are not formally trained in knowledge representation. We describe our approach for inconsistency monitoring in a large biology KB. We use a combination of anti-patterns that are indicative of poor modeling and inconsistencies due to underspecification. We draw the following lessons from this experience: (1) knowledge authoring must include an intermediate step between authoring and run time inference to identify errors and inconsistencies; (2) underspecification can ease knowledge encoding but requires appropriate user control; and (3) since real-life KBs are rarely consistent, a scheme to derive useful conclusions in spite of inconsistencies is essential.

The emerging paradigm of Precision Oncology 3.0 uses panomics and sophisticated methods of statistical reverse engineering to hypothesize the putative networks that drive a given patient's tumour, and to attack these drivers with combinations of targeted therapies. Here, we review a paradigm termed Rapid Learning Precision Oncology wherein every treatment event is considered as a probe that simultaneously treats the patient and provides an opportunity to validate and refine the models on which the treatment decisions are based. Implementation of Rapid Learning Precision Oncology requires overcoming a host of challenges that include developing analytical tools, capturing the information from each patient encounter and rapidly extrapolating it to other patients, coordinating many patient encounters to efficiently search for effective treatments, and overcoming economic, social and structural impediments, such as obtaining access to, and reimbursement for, investigational drugs.

Users routinely disclose personal information to obtain the benefits of Personalized Online Services. As a result, personal data is distributed across uncounted and unaccountable remote databases. Data mismanagement, as well as privacy and security flaws undermine individuals' control and privacy of their personal data. Yet revealing detailed private data does not necessarily yield useful service personalization; often this functionality is only modestly dependent upon the accuracy of user-supplied input. We demonstrate knowledge-based input generalization wherein systematically perturbed user data is supplied to a personalized service to gain forward privacy for the user, while retaining the utility of the service's results.