Page 6 - IOE_Winter 24
P. 6

RECENT


          DEVELOPMENTS IN AI











                                                               Intel, INSAIT, Linux Foundation, ML Commons, MOC
                                                               Alliance (operated by Boston University and Harvard
                                                               University), NASA, NSF, Oracle, Partnership on AI, Red
                                                               Hat, Sony Group, Stability AI, University of California,
                                                               Berkeley University of Illinois, University of Notre Dame,
                                                               University of Tokyo and Yale University.
                                                               The Alliance aims to empower a wide spectrum of
                                                               AI researchers, builders and adopters in order to
                                                               encourage collaboration and information sharing,
                                                               recognising the need for open and transparent
                                                               innovation.
          EU drafts World’s first ‘comprehensive’
          Laws on AI                                           The vision of the Alliance includes:
          The European Union has struck a provisional          •   Development of benchmarks and standards -
          agreement to enact the world’s first                     creating and deploying benchmarks, evaluation
          comprehensive laws to regulate artificial                standards, tools and resources for responsible AI
          intelligence.                                            development and use on a global scale.
          Proposals have been added to the AI Act that include   •   Advancement of an ecosystem of open
          measures such as critical safeguards for AI usage        foundation models - responsibly advancing
          within the EU and limitations on its adoption by law     diverse modalities in open foundation models,
          enforcement agencies. Consumers will also have           helping to address global challenges in areas
          rights to launch complaints as with GDPR.                such as climate and education.
          The European Parliament defines AI as ‘software      •   Enhancement of the AI hardware accelerator
          that can generate outputs, such as content or            ecosystem - encouraging contributions and
          decisions, that influence their environments based       adoption of essential software technology for AI
          on human-defined objectives’. Generative AI, such        hardware accelerators.
          as ChatGPT and DALL-E, which learn from vast data
                                                               •   Supporting global AI skills and research
          sets to create human-like content, would come
                                                                   - engaging the academic community in
          under this definition. ChatGPT can engage in text
                                                                   supporting researchers and students in AI model
          conversations, while DALL-E creates images from
                                                                   and tool research projects.
          textual instructions.
                                                               •   Promotion of safe and beneficial AI
          The new legislation is expected to come into force
                                                                   development - initiating projects that foster
          early in 2025.
                                                                   open and responsible AI development and
          [ Reference: McCane, K., Electronic Specifier, 11th.     hosting events to explore AI use cases.
          December 2023 ]
                                                               The AI Alliance aims to establish member-driven
                                                               working groups across all major topical areas along
          IBM and Meta launch AI Alliance                      with a governing board and technical oversight
          IBM and Meta have jointly launched the AI            committee.
          Alliance, uniting some 50 founding members and
                                                               [ Reference: West, P., Electronic Specifier, 5th.
          collaborators from various global sectors.           December 2023]
          The group includes AMD, Anyscale, CERN, Cerebras,
          Cleveland Clinic, Cornell University, Dartmouth, Dell
          Technologies, EPFL, ETH, Imperial College London,



                                                            6
   1   2   3   4   5   6   7   8   9   10   11