{"doi":"10.1111/jgs.18804","title":"Artificial intelligence and technology collaboratories: Empowering innovation in <scp>AI</scp> + <scp>AgeTech</scp>","abstract":"Launched in September 2021, the Artificial Intelligence and Technology Collaboratories (AITC) for Aging Research program is the newest of the seven centers programs funded by the National Institute on Aging (NIA), a part of the National Institutes of Health (NIH), and is dedicated to helping Americans live longer, healthier lives through the application of AI and emerging technologies (Figure 1). Three collaboratories, centered at Johns Hopkins University (JH AITC), the University of Massachusetts Amherst (MassAITC), and the University of Pennsylvania (PennAITech), and a Coordinating Center managed by Rose Li & Associates Inc. (RLA) comprise the “a2 Collective.” By committing more than $65 million over 5 years toward this program, NIA amplified the promise of technology, and AI in particular, to accelerate the development of solutions to help Americans, especially those with dementia, live where they most want: in their homes.1 NIA relies on proven tactics for the greatest impact: bringing together multiple disciplines to tackle societal challenges, cultivating timely data sharing where possible, identifying and enticing top talent in varied fields to focus on aging, and orchestrating a harmonized approach from the start, with national scope, to give the program staying power in a fast-evolving AgeTech ecosystem. NIA earmarked $40 million over 5 years for the a2 Collective to hold annual pilot award competitions. Our first three calls for applications drew nearly 700 applications from 45 states plus Washington, DC, Puerto Rico, and the U.S. Virgin Islands. A significant majority (71%) of the initial 60 awardees (from two competitions) include academic collaborators, reflecting the premium placed on research rigor (Figure 2). More than 40% of pilots are led by women, and about three-quarters relate to dementia. Specific examples include a machine learning-enabled, speech-based dementia screening tool for families and caregivers and a simple imaging and telemedicine system for remote eye (cataract) screening in disadvantaged populations by non-ophthalmologists. (See additional funded pilot descriptions at https://www.a2collective.ai/awardees.) Most pilot projects are developing or beta testing prototypes (60%) and evaluating prototypes in real-world conditions (30%); fewer are at the technology concept or discovery stage (7%) or pursuing commercial deployment or scaling up (3%). About 80% of funded pilots involve machine learning (ML) and significant proportions are developing user-facing software and platforms, wearables, smart household devices or utilities, and environmental sensors. Pilot awards to date were selected from an application pool constituted before the advent of widespread public engagement with powerful large language models (LLMs) at the end of 2022. Given the many LLMs now available, future calls are expected to see a dramatic increase in the number of projects building on generative AI and LLMs. The landscape is becoming saturated with technologies and platforms with overlapping functions. Comparative effectiveness research is increasingly needed to facilitate inevitable consolidation in the field. In the meantime, we need to continue to identify and support the most innovative technology and partnerships, invest in projects that take calculated risks and employ ethical AI design, pursue practical product-development goals, and demonstrate an understanding of the user experience and integration into clinical practice. Almost all awarded pilots are collecting human subjects data, but generally in small numbers. Giving more weight to projects that access large datasets (e.g., through academic research institutions, industry, or payers) and prioritizing the completeness of data will accelerate the development of more sophisticated, inclusive AI/ML methods. The pace of AI evolution may simply be too rapid for our pilot projects to reflect the current research frontier in real time. Reducing the time to project la","journal":"Journal of the American Geriatrics Society","year":2024,"id":464913,"datarank":0.0,"base_score":0.0,"endowment":0.0,"self_citation_contribution":0.0,"citation_network_contribution":0.0,"self_endowment_contribution":0.0,"citer_contribution":0.0,"corpus_percentile":null,"corpus_rank":null,"citation_count":4,"citer_count":0,"citers_with_citation_signal":0,"citers_with_endowment":0,"datacite_reuse_total":0,"is_dataset":false,"is_dataset_confidence":0.8971,"is_data_producer":false,"deposit_databanks":null,"is_oa":true,"file_count":0,"downloads":0,"has_version_chain":false,"published_date":"2024-01-01","fair_score":null,"fair_percentile":null,"algorithm_id":"datarank_citation_only_1hop_v6","ranking_scope":"data_only","authors":[{"id":264351,"name":"Peter Abadir","orcid":"0000-0002-8186-0066","position":1,"is_corresponding":false},{"id":103605,"name":"Alexis Battle","orcid":"0000-0002-5287-627X","position":2,"is_corresponding":false},{"id":19509,"name":"Rama Chellappa","orcid":"0000-0002-7638-1650","position":3,"is_corresponding":false},{"id":270945,"name":"Niteesh K. Choudhry","orcid":"0000-0001-7719-2248","position":4,"is_corresponding":false},{"id":412309,"name":"George Demiris","orcid":"0000-0002-6318-5829","position":5,"is_corresponding":false},{"id":554379,"name":"Deepak Ganesan","orcid":"0000-0003-2762-9194","position":6,"is_corresponding":false},{"id":286073,"name":"Jason Karlawish","orcid":"0000-0002-6880-2865","position":7,"is_corresponding":false},{"id":14812,"name":"Jason H. Moore","orcid":"0000-0002-5015-1099","position":8,"is_corresponding":false},{"id":57359,"name":"Jeremy Walston","orcid":"0000-0002-6965-2723","position":9,"is_corresponding":false},{"id":1049696,"name":"R. Li","orcid":"0000-0002-7438-1876","position":0,"is_corresponding":true}],"reference_count":1,"raw_metadata":null,"created_at":"2026-07-19T02:04:42.082401Z","pmid":"38407353","pmcid":null,"fwci":null,"citation_percentile":null,"influential_citations":0,"oa_status":null,"license":null,"views":0,"total_file_size_bytes":0,"version_count":0,"fair_f":null,"fair_a":null,"fair_i":null,"fair_r":null,"fair_zscore":null,"fair_rationale":null,"fair_model":null,"fair_agent_version":null,"fair_fulltext_source":null,"fair_has_llm":null,"fair_computed_at":null,"clinical_trials":[],"software_tools":[],"db_accessions":[],"linked_datasets":[],"topics":[]}