AI Eye: 25K traders bet on ChatGPT’s stock picks, AI sucks at dice throws, and more

31 May 2023

Cointelegraph By Andrew Fenton

Almost 25,000 investors have signed up to trade alongside ChatGPT as they follow the GPT Portfolio experiment from copy trading firm Autopilot.

The traders have bet a combined $14.7 million on the AI’s stock picks, which would average about $600 each if they all invested after signing up. They’re hoping to take even a small slice of a purported 500% return from one of the strategies backtested in academic research.

The GPT Portfolio gets the AI to analyze 10,000 news articles and 100 company reports to select 20 stocks for the $50,000 portfolio, updated each week. The initial picks included Berkshire Hathaway, Amazon, D.R. Horton and Davita Health. After two weeks, the portfolio is up around 2%, which is pretty much the same as the stock market. 

Interestingly the bottom five picks lost more in percentage terms than the top five gained — Dollar Tree lost 17% after it missed earnings — so it might be more sensible in future to only invest in GPT-4’s best five or 10 ideas, but we’ll see how it works out.

The smaller-scale ChatGPT Crypto Trader account is tweaking a similar strategy that gets GPT-4s advice on when to go long on Ethereum. He says it shows a profit of 11,000% backtested to August 2017, but in the real-world experiment since January, the portfolio is up by a third, while the Ethereum price has gained 60%.

It’s worth being careful using AI for trading, however. Crypto derivatives platform Bitget recently abandoned its experiment of using AI on the platform due to the potential for misinformation. A survey of its users found 80% of users had a negative experience with the AI, including false investment advice and other misinformation. 

Bitget Managing Director Gracy Chen says:

“AI tools, while robust and resourceful, lack the human touch necessary to interpret market nuances and trends accurately.”

The GPT Portfolio hopes that CNN is on the money. (Autopilot)

Are LLM’s stupid?

There are two extremes when it comes to thinking about large language models like GPT-4: some people maintain they are stupid mansplaining bots that confidently blurt out fake information, while others believe they will lead to artificial general intelligence (equivalent or better than human intelligence). Researchers from Microsoft published a 155-page paper called “Sparks of General Intelligence” back in March, arguing the latter was the case, apparently super impressed that the GPT was clever enough to work out how to stack a book, nine eggs and a laptop on top of each other.

Demis Hassabis, the co-founder of DeepMind, thinks the rate of progress is set to continue, meaning we may be just “a few years, maybe a decade away” from AGI. But robotics researcher and AI expert Rodney Brooks argues that large language models like ChatGPT are not going to lead to AGI. He says they don’t understand anything and can’t logically infer meaning.

“What the large language models are good at is saying what an answer should sound like, which is different from what an answer should be.”

Another skeptic is AI writer Brian Chau, who is writing a three-part series called “Diminishing returns in machine learning.” He argues that AI development is bumping up against hardware limitations and the extravagant cost of training larger models and is starting to slow. He puts the chances of AGI at less than 5% by 2043

ChatGPT’s loaded dice

One task that’s beyond ChatGPT is rolling a die and giving you a random number. Ask it to do so, and it’ll invariably roll a four on its first go. Ask it for a number between one and 10, and it’ll pick seven. Ask it for a number between one and 30, and it’ll pick 17. (Bard is similarly nonrandom) One Redditor got it to roll a die 50 times, with the bot returning “31 fours, 12 threes, 4 sixes, 3 fives, and no ones or twos.” 

What seems to be happening is the “random” numbers it produces are the ones that appear most frequently in its training data — because humans pick those same random numbers most often as well. In fact, the phenomenon of people picking seven when asked to choose a number between one and 10 is so well known, it’s even a magic trick/pick-up technique that The Game author Neil Strauss used to impress Britney Spears when he correctly “guessed” a number she’d chosen.

Read also


Features

Agents of Influence: He Who Controls The Blockchain, Controls The Cryptoverse


Features

Joe Lubin: The truth about ETH founders split and ‘Crypto Google’

AI job losses may inspire a revolution

Goldman Sachs suggests 300 million jobs will be lost to automation, while The World Economic Forum tips 83 million jobs will go in the next five years alone. Unlike previous waves of job losses as automation increased, these are likely to be concentrated among white-collar workers. James Marriot argues in The Times of London that this could lead to a violent revolution. He points to the French and Russian revolutions being led by disgruntled lawyers and journalists and highlights academic research across 78 countries that shows that middle earners are more likely to participate in political unrest than the poor.

He also highlights Peter Turchin’s theory of political unrest being linked to “elite overproduction” — universities churning out highly educated people who can’t get the jobs and status they believe they’re owed. Marriot writes:

“These are precisely the people who may be about to graduate into a world that has no need for them. If the worst happens, they are going to be angry.” 

Adobe Firefly

More than 70 million images were created in the first month after Adobe released the beta of its generative AI tool Firefly. The “generative fill” feature turns anyone into a Photoshop wiz; you simply circle an area of a pic then type in what you’d like to see — say a pool of reflective water. You can replace objects or backgrounds, and “expand” images to be as large as you like. 

Users have been having great fun expanding classic album covers and famous artworks.

Originally in a swimming pool, now in an ocean. Nevermind. (Twitter)The Clash’s London Calling Album (Twitter)Master of Cakes by Metallica (Twitter)Van Gogh’s The Starry Night in widescreen (Twitter)

The tech will be coming to Google’s Bard soon. Another new project called DragGAN enables users to manipulate 2D images as if they were in 3D, say by spinning a car around to see it from different angles, opening a dog’s mouth to show its teeth, and many more uses.  

Specialized AIs are the future

According to the trademark filing, JPMorgan is developing a ChatGPT service called IndexGPT to help its clients select investments. The news comes hot on the heels of BloombergGPT, which was trained on 40 years worth of financial data. Bloomberg claims it performs much better than GPT-4 in terms of financial analysis, although you’ll need to pay $24,000 a year to access it via Bloomberg Terminal, so we’ll take their word for it. 

Research does suggest, however, that AIs trained using domain-specific data in areas like science and medicine outperform general-purpose LLMs. Intel is hoping this fact will pay off with its 1 trillion parameter Aurora AI model, which is being trained on scientific texts.

The aim is to work out the biological processes related to cancer and other diseases and come up with drugs to treat them. The enormous potential of this approach was highlighted by the news this week that an AI system combed through 6,680 chemical compounds and found one that kills an antibiotic-resistant superbug called acinetobacter baumannii, potentially saving millions of lives over the coming decades.

Video of the week

Wes Anderson directs Avatar with the Peculiar Pandora Expedition.

AI creator Bilawal Sidhu transformed the creepy Tesla robot army video into an even creepier Terminator-inspired video in an hour.

  

You might also like

Trump supports bill to buy 1 million BTC — Senator Lummis  
Trump supports bill to buy 1 million BTC — Senator Lummis  

US President Donald Trump supports the BITCOIN Act and has a team of experts in the White House working to roll out landmark digital asset legislation in the coming weeks, according to Wyoming Senator Cynthia Lummis. Speaking at the Bitcoin 2025 conference in Las Vegas, Nevada, Lummis said she is bringing the BITCOIN ACT to the “attention of the American people and the world,” adding that, “President Trump supports the bill.”In March, Lummis reintroduced the BITCOIN Act — landmark legislation that directs the US government to acquire 1 million Bitcoin (BTC) over five years. The acquisitions would be financed using existing funds within the Federal Reserve System and the Treasury Department. As Cointelegraph reported, the Trump administration has reiterated the need to use “budget-neutral ways” to acquire Bitcoin without burdening taxpayers.Source: CryptoGoosAt the Bitcoin Conference, Lummis said the Trump administration has a team working on “digital asset issues,” including legislation on stablecoins, market structure and the Bitcoin Strategic Reserve.“They will probably roll out in that order,” she said.“The Senate Banking Committee has passed the stablecoin bill out of committee,” said Lummis, adding: “We’re getting close to being ready to have it on the floor. We’ve worked for untold hours with the minority party to satisfy them, and we should be voting on it the week before we get back from this break.”Related: Senator Lummis’ new BITCOIN Act allows US reserve to exceed 1M BitcoinGENIUS Act on stablecoins is “going to pass,” says White House crypto czarThe White House seems to be in alignment with Senator Lummis. Last week, Trump’s top crypto adviser, David Sacks, said the GENIUS stablecoin bill is “going to pass” the Senate with bipartisan support after clearing a key procedural vote on May 19.On May 19, the Senate voted 66 to 32 to advance debate on the GENIUS Bill. Source: US SenateGENIUS refers to the Guiding and Establishing National Innovation for US Stablecoins Act, possibly the most comprehensive federal push to establish a legal framework for dollar-pegged stablecoins.Stablecoins have become one of the most prominent use cases for blockchain technology, with some industry advocates arguing that they could help extend the US dollar’s dominance as the global reserve currency.Collateralized, dollar-backed stablecoins like Tether’s USDt (USDT) and Circle’s USDC (USDC) account for more than 85% of the $250 billion market, according to CoinMarketCap.Related: Former CFTC chair criticizes STABLE Act amid calls for urgent regulatory clarity

Growing BTC reserve requires Congressional legislation — VanEck exec  
Growing BTC reserve requires Congressional legislation — VanEck exec  

Building a permanent US strategic Bitcoin reserve would likely require targeted legislation rather than executive action, according to VanEck’s head of digital assets, Matthew Sigel. Speaking at Bitcoin 2025 in Las Vegas, Sigel said the most viable path forward may involve inserting Bitcoin mining incentives into the congressional budget reconciliation process.According to Sigel, the most effective path to growing a US strategic Bitcoin reserve would be through targeted amendments to congressional budget legislation. These could include tax credits for mining companies that use methane gas and other incentives aimed at encouraging miners to share a portion of their mined BTC with the federal government. He argued that such an approach would allow the reserve to grow organically over time. Sigel also highlighted the limitations of executive actions in achieving this goal:”The problem with executive action is that it’s going to prompt lawsuits. And anything over $100 million is going to get sued by the Elizabeth Warrens of the world. So, I would say start with something maybe in the Exchange Stabilization Fund for $100 million.”US President Donald Trump established the US Bitcoin Strategic Reserve through a March 7 executive order. According to the order, the US government can only acquire Bitcoin through budget-neutral strategies or asset forfeiture, prompting a range of different ideas on how to add to the government’s stockpile of nearly 200,000 BTC.From left to right, Alex Thorn, Matthew Sigel, Matthew Pines and Fred Thiel. Source: Turner Wright/CointelegraphRelated: Bitcoin’s new highs may have been driven by Japan bond market crisisLawmakers, officials pitch different ideas to grow strategic Bitcoin reserveWyoming Senator Cynthia Lummis, the US lawmaker who introduced legislation for a Bitcoin strategic reserve in July 2024, proposed converting a portion of the gold certificates held by the US Treasury to Bitcoin.Converting gold to Bitcoin would allow the US government to purchase more Bitcoin without incurring a cost to the taxpayer, Lummis said.Bo Hines, the executive director of the President’s Council of Advisers on Digital Assets, echoed the idea in March 2025.Hines called on the US Treasury to revalue its gold holdings, which are currently priced at just $42.22 per troy ounce, and convert a portion of those gains to Bitcoin. This strategy would also be budget-neutral, Hines said.The price of gold reached an all-time high of $3,500 per ounce in April but experienced a minor pullback to around $3,300 on May 27.Magazine: TradFi fans ignored Lyn Alden’s BTC tip — Now she says it’ll hit 7 figures: X Hall of Flame

ZKPs can prove I'm old enough without telling you my age  
ZKPs can prove I'm old enough without telling you my age  

Opinion by: Andre Omietanski, General Counsel, and Amal Ibraymi, Legal Counsel at Aztec LabsWhat if you could prove you’re over 18, without revealing your birthday, name, or anything else at all? Zero-knowledge proofs (ZKPs) make this hypothetical a reality and solve one of the key challenges online: verifying age without sacrificing privacy. The need for better age verification todayWe’re witnessing an uptick in laws being proposed restricting minors’ access to social media and the internet, including in Australia, Florida, and China. To protect minors from inappropriate adult content, platform owners and governments often walk a tightrope between inaction and overreach. For example, the state of Louisiana in the US recently enacted a law meant to block minors from viewing porn. Sites required users to upload an ID before viewing content. The Free Speech Coalition challenged the law as unconstitutional, making the case that it infringed on First Amendment rights. The lawsuit was eventually dismissed on procedural grounds. The reaction, however, highlights the dilemma facing policymakers and platforms: how to block minors without violating adults’ rights or creating new privacy risks.Traditional age verification failsCurrent age verification tools are either ineffective or invasive. Self-declaration is meaningless, since users can simply lie about their age. ID-based verification is overly invasive. No one should be required to upload their most sensitive documents, putting themselves at risk of data breaches and identity theft. Biometric solutions like fingerprints and face scans are convenient for users but raise important ethical, privacy, and security concerns. Biometric systems are not always accurate and may generate false positives and negatives. The irreversible nature of the data, which can’t be changed like a regular password can, is also less than ideal. Other methods, like behavioral tracking and AI-driven verification of browser patterns, are also problematic, using machine learning to analyze user interactions and identify patterns and anomalies, raising concerns of a surveillance culture.ZKPs as the privacy-preserving solutionZero-knowledge proofs present a compelling solution. Like a government ID provider, a trusted entity verifies the user’s age and generates a cryptographic proof confirming they are over the required age. Websites only need to check the proof, not the excess personal data, ensuring privacy while keeping minors at the gates. No centralized data storage is required, alleviating the burden on platforms such as Google, Meta, and WhatsApp and eliminating the risk of data breaches. Recent: How zero-knowledge proofs can make AI fairerAdopting and enforcing ZKPs at scaleZKPs aren’t a silver bullet. They can be complex to implement. The notion of “don’t trust, verify,” proven by indisputable mathematics, may cause some regulatory skepticism. Policymakers may hesitate to trust cryptographic proofs over visible ID verification. There are occasions when companies may need to disclose personal information to authorities, such as during an investigation into financial crimes or government inquiries. This would challenge ZKPs, whose very intention is for platforms not to hold this data in the first place.ZKPs also struggle with scalability and performance, being somewhat computationally intensive and tricky to program. Efficient implementation techniques are being explored, and breakthroughs, such as the Noir programming language, are making ZKPs more accessible to developers, driving the adoption of secure, privacy-first solutions. A safer, smarter future for age verificationGoogle’s move to adopt ZKPs for age verification is a promising signal that mainstream platforms are beginning to embrace privacy-preserving technologies. But to fully realize the potential of ZKPs, we need more than isolated solutions locked into proprietary ecosystems. Crypto-native wallets can go further. Open-source and permissionless blockchain-based systems offer interoperability, composability, and programmable identity. With a single proof, users can access a range of services across the open web — no need to start from scratch every time, or trust a single provider (Google) with their credentials.ZKPs flip the script on online identity — proving what matters, without exposing anything else. They protect user privacy, help platforms stay compliant, and block minors from restricted content, all without creating new honeypots of sensitive data.Google’s adoption of ZKPs shows mainstream momentum is building. But to truly transform digital identity, we must embrace crypto-native, decentralized systems that give users control over what they share and who they are online.In an era defined by surveillance, ZKPs offer a better path forward — one that’s secure, private, and built for the future.Opinion by: Andre Omietanski, General Counsel, and Amal Ibraymi, Legal Counsel at Aztec Labs.This article is for general information purposes and is not intended to be and should not be taken as legal or investment advice. The views, thoughts, and opinions expressed here are the author’s alone and do not necessarily reflect or represent the views and opinions of Cointelegraph.