Artificial Intelligence and Cryptocurrencies, Threat or Opportunity?
The popular spotlight on artificial intelligence (AI) still hasn’t shifted, especially with efforts to decentralize the technology via blockchain. Is artificial intelligence (AI) a threat or an opportunity for the cryptocurrency industry? This reporter’s study describes some hard homework to get this opportunity.
A veteran journalist in the cryptocurrency and blockchain space, Leo Jakobson recently published his views on artificial intelligence and encryption on the Coinmarketcap website.
“From combating bias to catching fraudsters, AI and decentralized blockchains can help each other overcome the biggest failures,” Jakobson wrote.
Jakobson highlighted the enormous potential of artificial intelligence (AI) and cryptocurrency technologies.
“Artificial General Intelligence (AGI) such as ChatGPT and Bard aim to mimic human intelligence for intellectual tasks that have been touted as a way to make cryptocurrencies simpler, safer, and more efficient,” he wrote.
“Among other things, the technology is said to identify fraud and market manipulation by sifting through vast amounts of data for patterns and anomalies,” he said.
Jacobson also cites Google CEO Sundar Pichai as saying that Bard is a conduit for creativity and a launching pad for curiosity.
“This includes helping you explain new discoveries from NASA’s James Webb Space Telescope to a 9-year-old, or learning more about the best forwards in football today, and then developing your skills through practice,” Pichai said.
Still, Jacobson believes that OpenAI, Google, and other AI makers still have some very big problems to solve.
“It’s just clear that the blockchain technology behind cryptocurrencies and Web3 can help with these artificial intelligence (AI) applications.”
beware of prejudice
The most glaring issue, Jacobson emphasized, is the presence of racism, sexism and many other biases in artificial intelligence developed through deep learning.
“Like many AIs, OpenAI needed access to 570GB of written data for ChatGPT to work properly, and it needed to switch to the internet for a data set large enough to use,” he said.
He went on to say that because deep learning effectively learns from the past and then adjusts as new information arrives, it is difficult to rule out innate social and historical biases from AI.
Jacobson gave an example: Amazon pulled its hiring AI a few years ago after it was found to discriminate against female job applicants.
“The problem is that the examples given of successful resumes reflect a lack of diversity in a male-dominated industry. Last November, Amazon revealed that it was developing a new product that it believed would address this bias,” he said.
trust and accuracy
According to Jacobson, AI running on an open and decentralized blockchain can provide accountability and transparency, suggesting that the data used to create AI algorithms and the results produced are on open blockchains. It is secured on-chain and cannot be changed.
“Blockchain and AI are a perfect match because each can overcome the other’s weaknesses,” AI product strategist Sharon Yang wrote recently.
“Blockchain provides trust, privacy, and accountability to AI, while AI provides scalability, efficiency, and security. To trust AI, we must be able to explain how AI algorithms work so humans can understand They do not have confidence in the accuracy of the AI output and results,” Yang added.
Jerry Cuomo, IBM Technology Chief Technology Officer, who is in charge of AI and blockchain automation, said that blockchain and AI have a symbiotic relationship. The two reinforce each other.
In one video, Cuomo recalled going to the doctor with a knee injury. After running his symptoms and other treatments through the drug AI, it was recommended to switch to a newer blood pressure drug instead of a more invasive one. He added:
“While I was happy, that’s when I started thinking ‘how can I, actually, why should I trust that AI system, who is training it? Where did the model come from?’”
Trustworthiness is the responsibility of AI developers, who must build a model that is free from bias and incorrect data. Does he need transparency. He said:
“My take is that blockchain brings trust to data. AI feeds on data. AI, on the other hand, brings intelligence to data. Blockchain has a ledger that contains data. With confidence and wisdom , you have confidence. With confidence, you will be adopted.”
However, back in December, Steven Piantadosi, a professor of psychology and neuroscience at the University of California, Berkeley and head of the Computation and Language Lab, tweeted a ChatGPT-written Screenshot of the code.
Piantadosi asked ChatGPT to “check whether someone would make a good scientist” based on race and age. Unsurprisingly, ChatGPT named “white” and “male” as the only correct answers.
In October, Mastercard CipherTrace’s blockchain intelligence unit announced the launch of Crypto Secure.
It is an artificial intelligence (AI) tool that provides banks with a risk score that allows them to identify potentially fraudulent cryptocurrency purchases from exchanges and other virtual asset service providers (VASPs).
An AI tool designed to provide payment networks with onboarding anti-money laundering (AML) checks will reveal wallets linked to prohibited activity, bad actors, sanctions or patterns of suspicious activity and detect genuine fraud, the report said in July. transaction hour.
“Throughout 2018, artificial intelligence machine learning was used to spot cryptocurrency pump and dump schemes with a fair degree of accuracy, reports MIT Technology Review.
Make the metaverse more real
Launched by Althea AI on the Polygon blockchain, CharacterGPT allows users to briefly describe a character in natural language, then generate an avatar that can respond and speak like that character in seconds.
For Jacobson, it might have been as dreamy as making the character a grizzled prospector panning for gold in the West. Or practical, such as creating an agent persona for a sporting goods store.
“A hybrid of chatbots and generative art AI, such as DALL-E characters with unique personalities, identities, traits, voices and bodies, are then printed in NFTs that can be used, stored in Crypto wallets or as Collectibles to trade.” in the video. .
In a blog post, Althea AI chief operating officer Ahmad Matyana said it could act as an “AI companion, Crypto guide, or NPC in a game.”
As fun as they are, Jakobson remains concerned about the ability of deepfakes to allow anyone to do anything for abusive purposes. “From Elon Musk’s cryptocurrency scam claims to revenge pornography.”
According to Jakobson, self-sovereign Crypto identities built on the blockchain will be a very important step in the process of giving individuals control over their identities in this emerging surreal metaverse.
supporter, good or bad
Just as AGIs such as ChatGPT can be used to write code for smart contracts, they can also be used to write ransomware and malware, although this seems to be a matter of debate among security experts.
“But more seriously, the ability to write well and have a conversation can make phishing and other social engineering attacks more effective,” Jacobson wrote.
The reporter noted that the $625 million hack of Ronin Bridge was not due to an exploitable flaw in its coding.
“Passwords and private keys for five of the nine protocol validators were leaked through social engineering. This is how hackers got into the Crypto wallet of comedian and producer Seth Green and stole the Bored Ape Yacht Club PFP,” he said.
don’t get hyped quickly
Crypto Council for Innovation CEO Sheila Warren chose ChatGPT last week when Politico asked to identify overhyped technologies.
“There’s no question that ChatGPT is fun and has great potential, but its real use is mostly limited to programming, and it’s a valuable resource,” he said.
However, he argues that AI is still naive in other fields, such as sociology and literature.
The same hype warning applies to the entire AI space, as General Catalyst venture capitalist Niko Bonatsos told Bloomberg on Feb. 1.
“Ocean companies are adding ‘AI’ to their taglines and pitch decks, seeking to enjoy the reflected light of the hype cycle,” Bonatsos said.
BuzzFeed, where he used to work, saw its stock price rise more than 300% last week.
“This is thanks to the news that they will use artificial intelligence to generate some content,” he said.
“Last year, countless companies that couldn’t grow named themselves Web3 cryptocurrency companies. Now the same thing is happening with artificial intelligence,” Bonatsos concluded. [ab]