Yonah Welker’s Post

Just around the time when the UNESCO ad hoc group is drafting a Recommendation on the Ethics of Neurotechnology where I have the opportunity to present as an observer, there is another good paper exploring neuro data protection and privacy which plays a role anywhere from EEG headsets to non-invasive rehabilitation techniques and NIBS Link - https://lnkd.in/eMthESBQ Authored by:  Jared Genser, Stephen D., and Rafael Yuste* The Neurorights Foundation #ai #ethics #policy #neuroethics #neurorights

United Nations University Institute in Macau AI Conference today, in discussion on how to use Synthetic Data to train AI models. Paper - https://lnkd.in/eSjsHCgJ (The Use of Synthetic Data to Train AI Models: Opportunities and Risks for Sustainable Development) My colleagues are there today. For the cognitive/sensory/health portfolios we oversee at the Massachusetts Institute of Technology European Commission, when created and deployed accurately, synthetic data can be a cost-efficient way to improve the accuracy of healthcare AI models in cases when obtaining real-world data is challenging. However, it's still not free from errors and may bring under-representation of certain patient groups (age, gender, ethnicities, comorbidities) or generalized procedures in data collection. with Auxane Boch Tshilidzi Marwala チリツィ マルワラ [丸笑] #ai #ethics #policy #un

Last year, we had the opportunity to contribute suggestions to the White House GenAI PCAST initiative. It was followed by the AI Bill and executive order, and now - by the new bipartisan AI bill, the Future of AI Innovation Act. Its statements come in parallel with the EU's AI Office, and innovation incentive package, including investments in language consortiums, oversight / safety institute, regulatory sandboxes for high-risk models, public exchange, etc In particular, the bill/act includes the following statements/details: ● Authorizes the NIST AI Safety Institute to develop AI standards; ● Creates new AI testbeds with national laboratories to evaluate AI models y (EU analog - regulatory sandboxes); ● Creates public competitions to empower private-sector AI innovation; ● Creates international alliances on AI standards, research and development. The link to the bill - https://lnkd.in/ewdmyxgD Our commentary - https://lnkd.in/e6dejiyD #ai #ethics #policy #aiact #aibill

Some time ago, I was invited to peer-review Deepmind's paper on the ethics of Advanced AI assistants but needed to do my evaluation work at MIT, so it seems like it was already released. The link to the paper - https://lnkd.in/e3NzSJFt Authored by Iason Gabriel Social assistants play a significant part in our work for cognitive, sensory, and physical impairments, health, educational and work applications. It's good to explore Google's take on its growing autonomy, potential, limitations and ethical concerns. #ai #ethics #policy

This year the United Nations Internet Governance Forum will be held in Riyadh, where I plan to present on accessibility-centered algorithms and digital policies, the multistakeholder digital cooperation protecting designated groups, including our suggestions for the Digital services / Online Act, EU-MENA algorithmic policy cooperation, outcomes of the safety treaties such Bletchley Declaration, geopolitics and diplomacy of generative AI, algorithms and ontologies behind it. It will follow my curation of the Global AI Summit for the good of humanity in Riyadh in 2022, which led to the signing of Unesco Ethics recommendation, and a year later - the launching of the ethics research center, cooperation on policies and ontologies. You may join us here - intgovforum.org #ai #ethics #policy #dsa #accessibility

Today will join as an observer to oversee the first meeting of the Ad Hoc expert group, drafting a Recommendation on the Ethics of Neurotechnology, UNESCO. You can join as an observer here - https://lnkd.in/e_hqsMNz It's critical for our extensive work, evaluating publicly-funded neurotechnology portfolios, research and technology transfer (US-EU-MENA bridge), deploying human/disability-centered adoption and oversight frameworks at the Massachusetts Institute of Technology and European Commission projects. The outcomes of our work and call to action were initially addressed to the member states and the parliament in 2021, to the AI Act and other groups in 2023. It includes solutions tackling cognitive and sensory impairment and disabilities, multistakeholder involvement and protection (patients, caregivers), modular nature of patient-support ecosystem (including data, models, systems oversight and privacy), challenges of accuracy in the area of diagnostics (EEG, MRI), treatment and drug research, including novel antidepressants and better understanding of neurotransmitters (currently, not higher than 80-90% for AI-powered systems), implementation of ethical and non-invasive rehabilitation techniques and NIBS. The Ad Hoc group involves 24 experts, including Roberto Andorno hervé Chneiweiss Najia El Abbadi Nita Farahany Riadh Gouider Judy Illes Ryota Kanai Yodchanan Wongsawat Lucia Machova Urdzikova Dimiter Prodanov #ai #ethics #policy #neuroethics #neurorights #disability #accessibility

Following our trilogy on disability-centered algorithms and policy, as a scientific evaluator I contributed to the SAPEA's "AI in science in the EU" (recommendation and evidence report review to the theme requested by the European Commission Margrethe Vestager). The scope includes the use of algorithms in R&D, science, avoiding historical and statistical distortions, overseeing effects on public systems and technologies, security and sustainability. The full report, resources and recommendations - lnkd.in/egjtK53T with Céline Tschirhart Scientific Advice Mechanism #ai #ethics #policy

Following our algorithmic ontology and trilogy of reports in health, education and labor, will join WomenTech Network Global Conference (on April 24). Will cover: ● how different AI systems and ontologies, including expressive and affective computing, may interact with different spectrums and impairments; ● historical and statistical contexts of errors and distortions; ● approaches to data, models, systems oversight; ● policies and frameworks to systems categorization, scenarios, risks and oversight. Session - https://lnkd.in/eeN_XDCk OECD repository MOOC - https://lnkd.in/epiQaQtk with Anna Radulovski #ai #ethics #policy