Tech
“We’re just scratching the surface” of cryptocurrencies and artificial intelligence – Microsoft executive – TradingView News
Microsoft believes that artificial intelligence (AI) is “the defining technology of our time” and has been at the forefront of both AI research and investment.
But that doesn’t mean the Seattle-based tech giant isn’t also paying close attention to the cryptoverse, including ways in which blockchain technology and artificial intelligence could one day support each other.
At the recent Cornell Blockchain Conference in New York, Yorke Rhodes, Microsoft’s director of digital transformation, blockchain and cloud supply chain, was asked how the company viewed this possible intersection of technologies.
“I think that as these two technologies progress, it is possible to create agents that combine the power of both. We’re just scratching the surface,” she said.
In a panel discussion titled “Crypto x AI,” Rhodes’ views were further explored by moderator Alex Lin, co-founder and general partner of Reforge, who asked: Will Microsoft have its own blockchain one day?
“There’s already a huge amount of cool stuff going on” in the cryptocurrency industry, including the open source community, Rhodes responded, so “why would we try to recreate something that [already] does he have that much investment?”
Rather, Microsoft’s focus today is more on optimizing existing technologies, such as layer 2 blockchain rollups. Rodi added:
«But we would [Microsoft] have you ever built an L1 blockchain? I do not believe.”
Cryptocurrencies are “well positioned”
Rhodes and Lin were joined on stage at the April 26 event at Cornell Tech by Neil DeSilva, chief financial officer of PayPal Digital Currencies; Matt Stephenson, head of research at Pantera; and Jasper Zhang, CEO and co-founder of Hyperbolic Labs.
Stephenson said that “cryptocurrencies are quite well positioned to be the ‘picks and shovels’ of some type of artificial intelligence,” particularly transformation and diffusion models. This is especially true given the likelihood of a “decentralized, multi-agent” AI future.
However, cryptocurrencies may have to play a secondary role to the advantage of artificial intelligence. Rhodes acknowledged that a “mass trend” like artificial intelligence tends to “suck a lot of the air out of the room” for other emerging technologies, including cryptocurrency/blockchain and Web3. Cointelegraph
“It’s a hot topic: the intersection or symbiosis between blockchain networks and artificial intelligence,” Lin commented. But it’s also susceptible to exaggerated claims, and it can sometimes be difficult to separate what’s exaggeration from what’s real.
There’s a lot of talk about decentralized graphics processing units (GPUs), for example, Lin continues, “but no one talks about latency,” the time it takes for data to transfer across a network.
Recent: Is Trump interested in cryptocurrencies? Bitcoin is the latest battleground in the US elections
Nowadays, AI requests or recommendations from a centralized network can be obtained quite quickly. However, due to “latency,” Lin said, decentralized networks will not produce these results as quickly.
However, Hyperbolic Labs’ Zhang didn’t think this would be a problem for decentralized networks like blockchains. “The inference is feasible,” he said.
Take for example a centralized network with a data center based in Texas that receives a request from a user in the UK. The data request “has to travel from the UK across the ocean to Texas and then back. So there is actually a huge delay,” Zhang said.
In comparison, with a reasonably sized decentralized network, a user could easily find a node in London to process the request locally, which would “effectively reduce communication overhead”.
In fact, Hyperbolic Labs recently launched an AI inference interface on the company’s decentralized network and achieved latency results comparable to centralized solutions, Zhang said.
A growing trend: small language models
Much of the conversation around AI these days focuses on large language models (LLMs) that require huge amounts of computing power. However, according to Rhodes, “there’s a lot going on in what you might call edge AI: getting smaller language models that actually work efficiently on phones and laptops.” This is:
“There’s a lot more compute available at the edge because models are getting smaller and smaller for specific workloads, [and] you can actually get a lot more out of it.
Microsoft has developed AI models in small languages that require less training data and computing power to develop and run, including the Phi-3 family of open models. His abilities are “really starting to come close to some of the big language models,” Rhodes said.
Regulators have artificial intelligence in their sights
Artificial intelligence is likely to come under intense scrutiny from regulators around the world in the coming years, just as cryptocurrencies have. What obstacles did the speakers foresee regarding government rules and regulations?
“I think the United States, in particular, is terrible at this [regulation]” Lin said, referring to the US Securities and Exchange Commission’s heavy-handed approach to regulating cryptocurrencies. “Now, [SEC Chair] Gensler came out and said we will regulate AI even more aggressively than blockchain digital assets.”
Fintech giant PayPal has launched its own dollar-pegged stablecoin, PayPal USD 
“I don’t think the United States is terrible at regulation,” DeSilva said. “Look at all the innovation here in the United States.”
Sure, it can sometimes be frustrating dealing with government authorities, but “regulators have a mission,” he explained: “Not to harm customers.” They are trying to protect consumers and there is nothing inherently wrong with that. Or, as she said from the stage:
“If you want your technology, your innovation, to be used by millions or billions of customers, you will have to interact with regulators.”
However, other jurisdictions, including the European Union, are becoming more welcoming to stablecoin issuers, and the United States needs to be aware of this. “If the United States doesn’t move faster, this is an advantage that will go up in smoke,” DeSilva acknowledged. “The United States has struggled to achieve the right level of urgency in the country.”
Finding the right amount of regulation could be even more difficult with AI. It will be difficult for regulators to manage the potential harm to consumers given the opacity of AI decision-making – the so-called black box problem – “and I think regulators are going to have to wrestle with that,” DeSilva added.
This opacity could actually provide an opportunity for blockchain technology with its transparency, immutability, and tracking capabilities. Lin said:
“You [can] blockchains present themselves as a kind of lord and savior, saying, ‘Regulators, we have this mechanism that can clear up the opacity associated with these black boxes.’”
Why AGI?
Lin concluded the session by asking the speakers to share their visions for the future of artificial intelligence. For example, will generalized artificial intelligence (AGI) become a reality within the next five to ten years? And what might we expect in the short term?
“In the near future, AI may be powerful enough for everyone to start using it,” Zhang predicted. “Every company will be an AI company, just as every company is an Internet company today.”
“I think in five to 10 years, AGI will become possible,” Zhang continued. “Look at how quickly AI models improve now, and with the help of decentralized infrastructure we can aggregate computation,” i.e. increase the overall volume of available GPUs, which should allow even smaller players to participate.
Elsewhere, zero-knowledge proofs (ZK proofs) “will disappear within three years,” Rhodes predicted, and will be replaced by fully homomorphic encryption (FHE), a technology that achieves zero trust “by unlocking the value of data across non-domain domains.” reliable without needing to decrypt it,” according to IBM.
Reader: DeFi conflict between MakerDAO and Aave reopens due to growing risk perceived by DAI
FHE will solve many privacy issues, Rhodes said, and could be particularly useful for the healthcare industry, including clinical trials involving sensitive personal data.
Rhodes, summing up, recalled the words of Ethan Mollick of the Wharton School: “The artificial intelligence you use today will be the worst version of artificial intelligence you have ever used.” The same could be said regarding ZK proofs and fully homomorphic encryption. Overall, IT structures that protect privacy will improve a lot, he said.
DeSilva has been working in technology and finance for several decades and has seen many concrete predictions come and go. “But I find this optimism [often] wins the day,” he told the gathering, adding:
“So my prediction is you [will] getting to AGI on time and that it is beneficial for people. This will require everyone’s work.”