Microsoft, a leading technology company, recognizes the importance of artificial intelligence (AI) and has been actively involved in AI research and investment. However, Microsoft is also interested in exploring the potential intersection of blockchain technology and AI.
Yorke Rhodes, Microsoft’s director for digital transformation, blockchain, and cloud supply chain, believes that as these two technologies progress, they can be combined to create powerful agents. He acknowledges that they are still in the early stages of exploring this potential.
During a panel discussion at the Cornell Blockchain Conference, Rhodes was asked if Microsoft plans to have its own blockchain. He responded by highlighting the interesting developments happening in the crypto space and the existing investments in open source projects. He emphasized the importance of optimizing existing technologies, such as layer-2 blockchain rollups.
The panel discussion further delved into the relationship between crypto and AI. Matt Stephenson, head of research at Pantera, suggested that crypto could play a significant role in supporting certain types of AI, particularly transformer and diffusion models. However, Rhodes acknowledged that AI has been dominating the conversation and overshadowing other emerging technologies like crypto and blockchain.
The discussion also touched upon the challenges of decentralized networks, such as latency. While centralized networks can provide quick results, decentralized networks may face delays due to data transfer across the network. However, Jasper Zhang, CEO of Hyperbolic Labs, believes that decentralized networks can still handle inference effectively, reducing communication overhead.
Another trend in AI is the development of small language models that run efficiently on devices like phones and laptops. Microsoft has been working on developing these small language AI models, such as its Phi-3 family of open models, which are approaching the capabilities of large language models.
The panelists also discussed the regulatory challenges that AI may face in the future. While the United States has been criticized for its approach to regulating cryptocurrencies, Neil DeSilva, CFO at PayPal Digital Currencies, defended the regulators’ mission to protect consumers. However, other jurisdictions like the European Union are becoming more welcoming to stablecoin issuers, and the U.S. needs to be mindful of that.
The opaqueness of AI’s decision-making process could present an opportunity for blockchain technology with its transparency and tracking capabilities. However, regulators will face difficulties in managing the potential harm to consumers due to AI’s black box problem.
Looking ahead, the panelists shared their visions for the future of AI. Zhang believes that AI will become so powerful that every company will be an AI company. He also predicts that AGI (artificial generalized intelligence) will become possible within the next five to 10 years. Rhodes mentioned the potential of fully homomorphic encryption (FHE) to solve privacy problems, particularly in the healthcare industry.
In conclusion, the panelists expressed optimism about the future of AI and its potential to improve with time. They believe that computing frameworks that prioritize privacy will continue to evolve and become more robust.