The text discusses a panel discussion at the Consensus 2022 conference on the topic of artificial intelligence (AI), decentralized infrastructure, regulation, and the future of AI. The panelists included Micah Zhang from Chainalysis, Chris Rhodes from DeepMind, and Dan DeSilva from Circle.
Microsoft holds the perspective that artificial intelligence (AI) represents a groundbreaking technology that significantly shapes our current era. The company has consistently spearheaded advancements in AI research and finance.
Seattle’s tech titan, Microsoft, may not be actively involved in the crypto world at present, but it certainly keeps a watchful eye on its developments. The potential synergy between blockchain technology and artificial intelligence is particularly intriguing to them.
During the Cornell Blockchain Conference held in New York City, attendees were curious to know Microsoft’s perspective on the potential convergence of these advanced technologies. Yorke Rhodes, Microsoft’s director for digital transformation, blockchain, and cloud supply chain, addressed this question.
“He believes that by combining the advances of these two technologies, we can develop powerful agents. We’ve only begun to explore their full potential.”
During the “Crypto x AI” panel discourse, Microsoft’s perspective on having their own blockchain was explored in greater depth by Alex Lin, Reforge’s co-founder and general partner, who posed the question.
In the dynamic world of cryptocurrency, there’s an abundance of intriguing projects and initiatives ongoing. Rhodes acknowledged this by stating, “We don’t need to add more interesting things to an already bustling scene.” Alternatively, “There’s plenty happening in crypto that’s captivating; why reinvent the wheel?”
As a Microsoft analyst, I would phrase it this way: Currently, our team’s primary goal is to enhance the performance of existing technologies, specifically focusing on implementing layer-2 blockchain rollups for better optimization. Rhodes further emphasized:
“But would we [Microsoft] ever build an L1 blockchain? I don’t think so.”
Crypto is “well positioned”
At the Cornell Tech event on April 26, Rhodes and Lin were accompanied on stage by Neil DeSilva, who serves as the CFO at PayPal’s Digital Currencies division; Matt Stephenson, the head of research at Pantera; and Jasper Zhang, the CEO and co-founder of Hyperbolic Labs.
As an analyst, I would interpret Stephenson’s perspective as follows: Crypto could serve as the essential infrastructure or foundational technology for certain types of advanced artificial intelligence (AI), specifically transformer and diffusion models. This is particularly relevant given the anticipated trend towards a decentralized, multiagent AI future.
Despite its significance, cryptocurrencies could potentially take a back seat to artificial intelligence’s (AI) dominance. According to Rhodes, the substantial wave that is AI often leaves less space for other emerging fields such as crypto/blockchain and Web3.
As a analyst, I’ve noticed that the intersection or symbiotic relationship between blockchain networks and artificial intelligence is a subject of great interest. However, it’s important to be mindful of the exaggerated claims surrounding this topic. At times, it can be challenging to discern what’s based in reality versus what’s merely hype.
As a researcher exploring the topic of decentralized graphics processing units (GPUs), I’ve noticed that while there is considerable discussion surrounding this technology, there seems to be a lack of focus on an essential aspect – latency. Latency refers to the time it takes for data to travel across a network.
Recent: Does Trump care about crypto? Bitcoin is latest battleground in US election
As an analyst, I’ve noticed that centralized AI networks are capable of providing responses or suggestions relatively swiftly due to their direct connection to a large data processing center. However, according to Lin, this advantage is outweighed in decentralized networks by the issue of latency. In simpler terms, with decentralized systems, the data might take longer to travel and be processed due to the absence of a single central hub.
As a analyst, I’ve come across the viewpoint of Zhang from Hyperbolic Labs who expressed doubt about the challenge posed to decentralized networks such as blockchains in terms of inference. Yet, in his perspective, “inference is feasible.”
A centralized network with a data center situated in Texas processes a user’s request originating from the United Kingdom. This data request necessitates a journey over the Atlantic Ocean to reach Texas and subsequently return, leading to a significant delay. (Zhang’s statement)
As a crypto investor, I can tell you that having a decentralized network of a reasonable size brings about several benefits. For instance, instead of relying on a distant server to process my requests, I could quickly locate a nearby node in London to handle it. This approach would significantly cut down on communication overhead and make the overall transaction process more efficient for me.
Recently, Hyperbolic Labs introduced an AI inference interface on their decentralized system, delivering latency outcomes similar to those obtained from centralized alternatives, according to Zhang.
A growing trend: Small language models
A significant portion of the discourse surrounding artificial intelligence today revolves around large language models (LLMs), which demand substantial computational resources. Nevertheless, as Rhodes points out, there’s also a growing trend towards edge AI, where smaller language models are developed to operate effectively on mobile devices and laptops.
“There is much more compute available at the edge because the models are getting smaller for specific workloads, [and] you can actually take a lot more advantage of that.”
Microsoft is working on creating compact language AI models from its Phi-3 series of open models. These models need less data for training and fewer computational resources to operate. According to Rhodes, their abilities are becoming quite comparable to those of larger language models.
Regulators have AI in their sights
In the next few years, AI may face significant examination from regulatory bodies worldwide, similar to the scrutiny cryptocurrencies have experienced. Which challenges did the panel predict regarding governmental rules and regulations?
“According to Lin’s perspective, the United States struggles with regulation. He mentioned the strict regulatory measures adopted by the Securities and Exchange Commission (SEC) towards cryptocurrencies as an example. Moreover, SEC Chair Gensler has recently announced plans for increased regulatory oversight of AI technology, surpassing the current regulations for digital assets.”
Seven months ago, PayPal, a leading fintech company, introduced its own dollar-backed digital currency named PayPal USD (PYUSD). With this development, Lin was curious to hear DeSilva’s perspective on the regulatory landscape in the United States.
“According to DeSilva, the US isn’t inept when it comes to regulations. Witness the abundance of inventive ideas and advancements originating from American soil.”
As an analyst, I understand the challenges of interacting with regulatory bodies. However, it’s crucial to keep in mind that their primary role is to ensure consumer safety. They’re working diligently to prevent any potential harm from reaching customers. There’s nothing sinister about that intention.
“If you want your technology, your innovation, to be used by millions or billions of customers, you’re going to have to engage with regulators.”
Despite other regions, such as the European Union, growing more receptive to stablecoin issuers, the United States remains cautious. However, this hesitancy could result in lost opportunities if action isn’t taken swiftly. “If we don’t act quickly, that advantage will slip away,” DeSilva conceded. “The U.S. has faced challenges in establishing the necessary urgency regarding this issue.”
Regulating the appropriate level of control over artificial intelligence (AI) may prove to be a complex task due to AI’s intricate decision-making mechanisms, which are often described as opaque or “black boxes.” Regulators might find it challenging to prevent potential consumer harm given this lack of transparency. As DeSilva noted, “I believe they will face significant struggles in dealing with this issue.”
The lack of clarity surrounding that situation might instead present a valuable chance for blockchain tech, given its inherent transparency, unchangeable records, and robust tracking features. (Lin stated)
“You [can] have blockchains coming in as a kind of lord and savior, saying: ‘Regulators, we have this mechanism that can clear up the opacity associated with these black boxes.’”
Wherefore AGI?
At the end of the discussion, Lin posed a question to the panel members regarding the feasibility of achieving Artificial Generalized Intelligence (AGI) within the upcoming five to ten years. Furthermore, they were invited to describe potential developments in AI that could materialize in the nearer future.
As a crypto investor, I firmly believe that we’re on the brink of a technological revolution where artificial intelligence (AI) will become a game-changer for businesses and individuals alike. It won’t be long before AI surpasses its current capabilities, making it an essential tool for everyone, just as the internet is today. Every company, irrespective of their industry or size, will need to embrace AI to remain competitive in tomorrow’s market.
In the next five to ten years, I believe AGI (Artificial General Intelligence) will become a reality. Observe how swiftly AI models advance currently. By utilizing decentralized infrastructure, we can merge computational resources, significantly increasing the quantity of accessible GPUs. This expansion should not only expedite progress for larger entities but also enable smaller participants to join the endeavor.
As a crypto investor, I believe zero-knowledge proofs (ZK-proofs) will be surpassed in the next three years. This prediction comes from my understanding that fully homomorphic encryption (FHE) will take over. FHE is a game-changing technology that allows the computation of encrypted data without decrypting it first, ensuring zero trust even on untrusted domains.
Reader: MakerDAO and Aave’s DeFi conflict reopens over DAI’s perceived risk growth
As a privacy analyst, I can affirm that Fully Homomorphic Encryption (FHE) holds immense promise in addressing numerous privacy concerns. Its potential applications are particularly noteworthy within the healthcare sector. For instance, FHE could significantly enhance the security and confidentiality of clinical trials dealing with sensitive personal data.
Rhodes, recounting Mollick’s words from Wharton School, stated, “The AI technologies you currently employ represent the most rudimentary form of AI you will ever encounter.” Similarly, ZK-proofs and fully homomorphic encryption are presently suboptimal. However, according to Rhodes, advancements in computing frameworks designed to uphold privacy will significantly improve in the future.
DeSilva, with his extensive experience spanning tech and finance industries for several decades, has witnessed numerous specific forecasts unfold. Yet, he shared with the audience, “I usually find that optimism prevails.”
“So my prediction is you [will] get to AGI in time, and that it’s a beneficial thing for people. That will take everybody’s work.”
Read More
- DOGS PREDICTION. DOGS cryptocurrency
- SQR PREDICTION. SQR cryptocurrency
- LDO PREDICTION. LDO cryptocurrency
- METIS PREDICTION. METIS cryptocurrency
- UXLINK PREDICTION. UXLINK cryptocurrency
- JASMY PREDICTION. JASMY cryptocurrency
- STG PREDICTION. STG cryptocurrency
- CLOUD PREDICTION. CLOUD cryptocurrency
- QUINT PREDICTION. QUINT cryptocurrency
- KNINE PREDICTION. KNINE cryptocurrency
2024-05-16 16:54