300+ international AI experts from tech giants including Meta and Google DeepMind exploring topics such as ethical considerations in AI ownership, innovations in sustainable AI computing, and compute power
Abu Dhabi, UAE – 26 November 2024: The Open-Source AI Summit Abu Dhabi, hosted by the Technology Innovative Institute (TII), a global applied science research center, has kicked off with critical conversations that will shape the global AI agenda. The summit, taking place at the St. Regis Saadiyat Island today and tomorrow with 300+ in attendance, converges with growing international focus on the pull and push between open- and closed-source AI.
There are two critical choices when referring to AI,” said H.E. Faisal Al Bannai, the Secretary General of the Advanced Technology Research Council and Adviser to the UAE President. “You can leverage a closed-source AI model that belongs to a company. They control it, including the data you give it. Innovation begins and ends with them.
“Alternatively, you can leverage an open-source model that grows within the community. We innovate together, and anyone, anywhere, can access and build on it. If AI is going to be part of the fabric of our society—and it will—countries, companies, and individuals need to decide who controls it. Opening Falcon AI models was TII’s contribution to the world.’”
Dr. Najwa Aaraj, the Chief Executive Officer of TII, who opened the summit, said, “The Open-Source AI Summit Abu Dhabi is a defining moment for global AI discourse. Falcon, like other open-source models, unites scientists, developers, and innovators to accelerate technological advancements as a catalyst for global change. We look forward to seeing the impact it will continue to have, particularly as we continue our work with the Falcon Foundation.”
The summit agenda continued with discussions from renowned speakers. including Dr. Belgacem Haba, the Vice President of the Adeia Corporation in the U.S., who spoke about the challenges that AI has created in the semiconductor manufacturing industry.
Prof. Philip Torr, a Professor and Chief Scientific Advisor at Oxford University in the U.K, spoke about who should own AI, touching on potential drawbacks as well as regulation, and arguing that in the long-term, the benefits of open-source AI outweigh the risks.
Dr. Hakim Hacid, the Chief Researcher of TII’s AI Research Center, said, “We believe open-source AI is the way forward, but it is far from simple—there are challenges and questions about control, policy, compute power, and hardware that we need to address. That is why we’re convening so many global experts at this summit and will continue to do so in future years in partnership with the Falcon Foundation. These conversations are crucial.”
Later speakers will include Dr. Natalia Vassilieva, the Vice President and Field CTO of Cerebras Systems in the U.S., Dr. June Paik, the Founder and CEO of FuriosaAI in the U.S., Dr. Armand Joulin, a Research Director at Google DeepMind in France, and Dr. Michal Valko, the Principal Llama Engineer at Meta Paris in France. They will speak about degrees of AI openness, sustainable AI computing, creating more compact LLMs, leveraging foundation models for trustworthy algorithms, and more.
TII’s Dr. Jingwei Zuo will speak about Falcon Mamba, the first State Space Language Model, built using entirely new architecture and launched earlier this year.
The summit will culminate in a panel led by TII on their vision for open-source AI. TII’s Falcon AI LLM series has garnered global recognition.
The series began with the launch of Falcon 40B, the UAE’s first open-source LLM, in May 2023. Since then, subsequent Falcon models have consistently ranked among the top open-source AI models globally, as validated by the independent industry leaderboard, Hugging Face. A new model of Falcon is expected by the end of 2024.
