Sam Altman Reveals: ‘Abu Dhabi Talked AI Before It Was Cool’

On June 6, 2023, a remarkable discussion unfolded in the Abu Dhabi Global Market (ADGM), hosted by Hub71, featuring leading AI visionaries and emerging tech entrepreneurs. At the core of the discourse was AI’s profound influence on startups and how it would bridge the generational divide and revolutionize work.

OpenAI’s CEO, Sam Altman, has been globe-trotting, spreading the word about the transformative power of AI and the significant role the UAE can play in guiding its future. Altman, in his latest pit stop in Abu Dhabi, made some notable remarks about the company’s ground-breaking AI product, ChatGPT, and its potential impact on global innovation.

Sam Altman, praised Abu Dhabi’s foresight in recognizing the potential of AI, stating that the city “has been talking about AI since before it was cool."

Altman is ambitiously making strides to promote safer AI use. He captivated audiences in Abu Dhabi during an open forum hosted by Hub71, the government-supported startup ecosystem. Altman shined a light on the necessity of complying with international laws, including proposed EU regulations on AI tools, and maintained that OpenAI was committed to operating within the legal boundaries.

In a refreshing clarification of his previous statements, Altman clarified, “We did not threaten to leave the EU. What we said is if we can’t comply with the law, then we won’t break it.” He also expressed his approval of European regulation, seeing it as a step in the right direction.

The vision of a global coalition to oversee the development and use of AI, as advocated by Omar Al Olama, the UAE’s Minister of State for AI, found a vocal supporter in Altman. He suggested a similar framework to that of the International Atomic Energy Agency, highlighting the need for worldwide cooperation in promoting the safe use of AI technologies.

Beyond the discussions, the prospect of a partnership with Mohamed bin Zayed University of Artificial Intelligence shows the growing influence of AI in shaping the future. Eric Xing, the president of the university, expressed his anticipation of working closely with Altman on pioneering “more powerful, useful and safe AI.”

In an era where AI is evolving rapidly and taking center stage, Altman’s approach to keeping humans involved in its development resonates strongly. As he put it, “It would be a mistake to take humans out of the loop. The future of humanity is decided by humanity, and that is an active decision we can make.”

While staying realistic about potential risks of AI misuse, Altman conveyed an optimistic future outlook, hinting at a world amplified by AI technologies. Although he didn’t reveal what future GPT’s like GPT-10 might look like, he said it would be a “giant multiplier on human ability and productivity.”

In essence, Sam Altman’s visit to Abu Dhabi is a testament to the UAE’s commitment to AI and the significant strides the nation is making in this field. It marks an exciting era in which AI is increasingly viewed not just as a tool, but as a catalyst for progress and societal enhancement. In Altman’s own words, “We will look back at this period of 2023 and say, ‘We didn’t have very good lives then, relative to what’s possible now.'”

Mohammed Sinan from Litz-Pattner prompted an interesting discussion when he asked Altman about the impact of large language models on small-scale researchers with limited funding and job security for software engineers.

Altman explained that while the race to scale up large language models might be out of reach for most academics, there are myriad other avenues for contribution. “There’s so much more we need to do. We have great partnerships with a lot of academics doing work on top of the models or work that feeds into the models. We desperately need new efficiency gains, better architectures, algorithms,” he said.

Regarding job security for software engineers, Altman dismissed concerns about AI replacing them. He argued that the rise of tools like GPT and Copilot has actually increased productivity and the demand for sophisticated code.

A key point that emerged from the conversation was the ethical implications of artificial intelligence. When asked how AI systems could be designed to reflect shared human values, Altman suggested a unique approach: to learn ethics rather than defining them in code. “Maybe you can learn it… And then from all of that, the system can learn the collective moral preferences of humanity. I think that’s pretty exciting,” he suggested.

Altman further addressed concerns about AI’s potential to become ‘out of control’ or ‘irreversible,’ a point termed as the ‘singularity’. In response to college student question, he emphasized the need for humans to remain in control of AI development and deployment.

Entrepreneurship, a critical skill for the rapidly changing landscape of AI, was another important topic of discussion. In response to Dr. Ifat Turbiner’s question on essential qualities for entrepreneurs, Altman emphasized resilience, adaptability, effective communication, long-term orientation, and productivity per unit time.

In a rapidly evolving technological landscape, it’s conversations like these that provide valuable insight into future trends, ethical considerations, and implications for our society. Such conversations to continue globally, creating a framework for the responsible use and development of AI technologies.

Altman addressed Simeon Kerr from the Financial Times’s question about AI’s role in the region. He expressed optimism about the commitment and desire for AI discussion in Abu Dhabi and the region’s potential to play a significant role in shaping global AI regulation and policy.

Former Y Combinator president and current CEO of OpenAI, Altman, emphasized AI’s potential to spark the greatest startup boom since the internet. According to Altman, AI is “the thing that is going to matter in the next decade” and recommended to “go all in on AI”.

Zohair Khaliq, co-founder of Jalebi, a restaurant tech company, directed the conversation towards the challenges brought by generational gaps. With a young son and an elderly father separated by a wide geographical and technological chasm, Khaliq wondered how AI could facilitate communication between the generations. It was a personal inquiry with far-reaching implications for families around the world grappling with the accelerating pace of technological change, a concept known as Future Shock.

Altman responded optimistically, providing a brief history of computer interfaces and their evolving simplicity. From punch cards to natural language interfaces, he suggested that the trajectory of technology has always been towards making technology more accessible for everyone. He shared his hope that AI, by leveraging natural language interfaces, would help bridge the generational divide, stating that “Everybody knows how to talk, almost everybody knows how to text, and we’re now at a kind of… the computer that sci-fi always predicted”.

Greg Armand, who runs a film production house, brought the discussion to the societal implications of AI. Referring to Altman’s blog post predicting super-intelligent AGI in the next 10 years, Armand questioned the likelihood of a universal basic income (UBI) being adopted due to the job displacement AI might cause. Altman acknowledged the potential of a UBI but also emphasized that work provides fulfillment and purpose beyond mere financial gain. Altman showed optimism that new types of work would arise with the advent of more advanced tools, ensuring that people could still find purposeful employment.

Salim Al-Haddad, an AI-focused student at Khalifa University and an early-stage fund participant in Silicon Valley, asked Altman about the prospects of a smaller team (even a one-person team) creating a billion-dollar company leveraging AI. Altman confirmed the possibility, hinting at the transformative power of AI to multiply productivity.

Responding to a query from Dana of CNN Business Arabia about OpenAI’s stance on EU AI regulations, Altman clarified that OpenAI was eager to comply with European regulations and looked forward to engaging with content creators about copyright issues in the new AI context.

Kashif, the founder of Payco, talked about their AI platform called Phantom that assists startups and SMBs in streamlining their business processes. Kashif sought advice on scaling their initiative across different geographies. Altman encouraged leveraging LLMs (like GPT-4) for content generation and suggested reaching out if they encountered difficulties, indicating the ongoing development and iteration in AI research.

Future GPTs to GPT-10

In his exchange with Dr. Ketaby from Abu Dhabi Executive Office, Altman cast a visionary glance into the future, musing on the prospect of GPT-10. He emphasized its potential to become “an extraordinary thing,” a “giant multiplier on human ability and productivity”. However, Altman was candid about the challenges of making precise predictions. “Predicting that for how it is… we’re talking something that’s more than a decade out,” he admitted.

Responding to Lee from HGM FinTech regarding the problem of hallucinations in large language models (LLMs), Altman expressed optimism. He affirmed OpenAI’s commitment to reducing these errors and assured that it was a priority research area for the team, emphasizing their ambition to ensure better usability of these models across regulated industries.

The conversation took an interesting turn when Louis from Venom Blockchain Foundation questioned about adding memory and temporal understanding to LLMs, enabling AGI to make better-informed decisions. Although Altman didn’t provide a timeline, he acknowledged the importance of the question and assured that such a capability would be a future research focus.

Sam Altman also expressed enthusiasm for collaboration when Jose, a technology creator focusing on disability accessibility, shared the impact of their tools in conjunction with OpenAI’s ChatGPT. Highlighting the efforts OpenAI is making to enhance accessibility, Altman acknowledged that “even with our current systems, people with disabilities… are able to use our speech recognition system that may not be able to use others.”

Additionally, Altman shared his views on AI’s transformative role in education. He envisions AI as a personalized tutor, capable of understanding individual learning styles and tailoring education to the needs of students across majors. “We can look in the future, but I think we already see what’s happening,” he said.

Addressing a question on AI safety measures, specifically the alignment problem, Altman explained how they’ve used Reinforcement Learning from Human Feedback (RLHF) extensively in GPT-4. However, he cautioned against overconfidence, emphasizing the need for continuous innovation as AI technology advances. The CEO emphasized OpenAI’s focus on developing new techniques, affirming his excitement about their current projects and the future of AI safety measures.

The discussion reached its zenith when Sam Altman was confronted with a question about his decision to not hold equity in OpenAI. In response, he shared his vision, stating that his fulfillment in working on OpenAI outweighed the monetary benefits. “The value I get personally from doing OpenAI so much eclipses what more money would do for me,” he reflected.

OpenAI’s Sam Altman in Abu Dhabi: The Future of AI and its Global Impact

When asked about OpenAI’s plans for an initial public offering (IPO), Altman dismissed the idea, citing the organization’s unconventional structure and its desire to avoid undue pressure from traditional public market investors. He emphasized the importance of having the freedom to make unconventional decisions, especially when the development of super intelligence comes into play.

Abdallah Lechalam from Emirates Sports Group raised a thought-provoking question about how technologies like AI, Web 3.0, Blockchain, and the Internet of Things (IoT) might synergistically disrupt existing systems. Altman envisioned Blockchain playing a significant role in validating content created by AI, as well as in the distribution of wealth and governance of AI systems. Regarding AI integration in smartphones, Altman confidently predicted that within a few years, a considerable portion of smartphone usage would resemble how we interact with chatbots like GPT.

Vera from LAMSA expressed concerns about the potentially catastrophic effects of AI on humanity if not properly regulated. Altman acknowledged the gravity of these concerns, praising the increasingly widespread attention AI is receiving. He stressed the importance of a global conversation and policy maker involvement in ensuring AI safety and responsibility.

Naima Felassi from Mubadala queried about the ideal model for AI regulation. Altman responded by proposing a multi-tiered model, suggesting that a global regulatory framework for potentially global consequences should exist, coupled with local determinations within broad global boundaries for smaller systems. Altman emphasized the need for global regulation for global effects, similar to how nuclear material is regulated. However, he hoped that there wouldn’t be excessive regulation on smaller systems with manageable risks, allowing them to develop and thrive.

Dima from Bayzat asked about the competitive advantage that companies can gain from integrating AI. Altman compared this phase of AI integration to the early days of the iPhone App Store, suggesting that being ‘AI-powered’ will soon become an expectation rather than a differentiator. The key to competitive advantage will still lie in a company’s core strategy and value proposition, not merely in AI adoption.

Willie Gomez, ambassador from Guatemala, inquired about biases within AI systems. Altman argued that AI could help reduce bias since AI lacks the inherent psychological flaws present in humans. He did, however, express concern about a potentially biased value system for AI if the alignment data primarily comes from a small number of viewpoints.

Ahmed Haghazi from Emirates Post asked about OpenAI’s future in robotics and potential integration with Tesla bot or Boston Dynamics. Altman revealed that OpenAI once pursued robotics but had to put it on hold to focus on language model development due to resource constraints.

Altman in Abu Dhabi & the Future of AI

This dialogue underscored the urgent need for a global, collaborative approach to AI, thereby ensuring a future where AI works for the benefit of all. The Emirates, and in particular Abu Dhabi, have already exhibited their commitment to this global conversation. The UAE launched the world’s first AI university, reaffirming their dedication to fostering innovation and AI-focused education. It’s essential for such conversations to continue globally, providing the groundwork for the responsible use and development of AI technologies.

The conversation at ADGM highlighted the immense potential of AI, not only as a business asset but also as a tool for social and personal interaction. As Altman suggested, we may be on the cusp of the most significant startup boom since the internet, facilitated by advancements in AI.

On a concluding note, Altman reaffirmed OpenAI’s commitment to realizing AGI’s potential broadly, aiming to reflect the expansive totality of human diversity. The event underscored the importance of ongoing dialogue and international cooperation in shaping the future of AI, with Altman showcasing the vision of OpenAI in this exciting journey towards AGI.

One thing is certain: those who wish to be at the forefront must be ready to embrace the AI revolution. With Abu Dhabi’s pioneering efforts and OpenAI’s ambitious vision, the AI landscape is set for transformative changes. The dialogue at ADGM was a valuable deep-dive into the world of AI, laying the groundwork for a future where AI serves us all.