Co-Developing Accessible Advanced AI
Download PDF
Co-Developing Accessible Advanced AI
Use the future to build the present
Co-Developing Accessible Advanced AI
Comment
Stakeholder Type
5.5SyntheticBiology5.4Science ofthe Originsof Life5.3FutureEconomics5.2Future ofEducation5.1ComplexSystemsScience4.4Democracy-affirming Technologies4.1Science-basedDiplomacy4.2Advances inScience Diplomacy4.3Digital Technologiesand Conflict3.7InfectiousDiseases3.6Solar RadiationModification3.5OceanStewardship3.4SpaceResources3.3Future FoodSystems3.2WorldSimulation3.1Decarbonisation2.6FutureTherapeutics2.5Organoids2.4ConsciousnessAugmentation2.3RadicalHealthExtension2.2HumanApplicationsof GeneticEngineering2.1CognitiveEnhancement1.6CollectiveIntelligence1.5AugmentedReality1.4BiologicalComputing1.3Brain-inspiredComputing1.2QuantumTechnologies1.1AdvancedAIHIGHEST ANTICIPATIONPOTENTIAL
5.5SyntheticBiology5.4Science ofthe Originsof Life5.3FutureEconomics5.2Future ofEducation5.1ComplexSystemsScience4.4Democracy-affirming Technologies4.1Science-basedDiplomacy4.2Advances inScience Diplomacy4.3Digital Technologiesand Conflict3.7InfectiousDiseases3.6Solar RadiationModification3.5OceanStewardship3.4SpaceResources3.3Future FoodSystems3.2WorldSimulation3.1Decarbonisation2.6FutureTherapeutics2.5Organoids2.4ConsciousnessAugmentation2.3RadicalHealthExtension2.2HumanApplicationsof GeneticEngineering2.1CognitiveEnhancement1.6CollectiveIntelligence1.5AugmentedReality1.4BiologicalComputing1.3Brain-inspiredComputing1.2QuantumTechnologies1.1AdvancedAIHIGHEST ANTICIPATIONPOTENTIAL

Opportunities:

Co-Developing Accessible Advanced AI

    There are 56 artificial intelligence (AI) startups worth over $1 billion today. That is a testament to the enormous power of deep learning, which has found transformative applications in everything from finance to healthcare. These approaches require huge amounts of data and computational power, however, which means that advances are increasingly driven by a handful of large companies and governments.

    We are about to enter a “third wave” of AI that will imbue machines with “common sense” and reasoning capabilities, allowing much broader deployment, and increasing the breadth and depth of human-machine interactions. That makes it crucial that these advances are not shaped by narrow interests and that everyone can take part in the development of advanced AI and benefit from its use.

    • What will the next generation of AI look like and how should we best prepare for it?
    • What priorities should inform the next stage of AI development?
    • How will advanced AI be able to address global challenges differently than today's technology?
    • What can we do to avoid “AI nationalism” and ensure broad access to the technology and applications developed on the basis of advanced AI?

    Takeaway messages

    The rise of AI is moving from data input to context and experience.
    Deep scepticism remains with AI algorithms, data sets and technology, which experts point out is culturally constructed. More regulation may be needed to prevent bias.
    Open source data and data-sharing in a fair and equitable way is essential to create new possibilities and solve scientific and technological problems.
    Science and technology organisations should incorporate people and strategies that reflect diverse sets of interests from the start when developing projects, not as an afterthought.
    Few scientific areas have not been impacted or completely revamped by the “pervasiveness” of advanced AI, and many things can go wrong if improperly used.
    Digital technologies and AI condition access to the world for an ever-larger group of people, making inclusiveness, representativeness and cultural biases ever more important.
    Inclusivity means building a table for everyone to gather around in the first place, not just adding seats.
    More reliance on 21^st century governance --- such as “soft law” that is not always enforceable but can be applied quickly through standards, practices, codes of conduct and insurance --- might be useful.