Connect with us

Published

on

David Silver, leader of the reinforcement learning research group at DeepMind, being awarded an honorary “ninth dan” professional ranking for AlphaGo.
JUNG YEON-JE | AFP | Getty Images

Computer scientists are questioning whether DeepMind, the Alphabet-owned U.K. firm that’s widely regarded as one of the world’s premier AI labs, will ever be able to make machines with the kind of “general” intelligence seen in humans and animals.

In its quest for artificial general intelligence, which is sometimes called human-level AI, DeepMind is focusing a chunk of its efforts on an approach called “reinforcement learning.”

This involves programming an AI to take certain actions in order to maximize its chance of earning a reward in a certain situation. In other words, the algorithm “learns” to complete a task by seeking out these preprogrammed rewards. The technique has been successfully used to train AI models how to play (and excel at) games like Go and chess. But they remain relatively dumb, or “narrow.” DeepMind’s famous AlphaGo AI can’t draw a stickman or tell the difference between a cat and a rabbit, for example, while a seven-year-old can.

Despite this, DeepMind, which was acquired by Google in 2014 for around $600 million, believes that AI systems underpinned by reinforcement learning could theoretically grow and learn so much that they break the theoretical barrier to AGI without any new technological developments.

Researchers at the company, which has grown to around 1,000 people under Alphabet’s ownership, argued in a paper submitted to the peer-reviewed Artificial Intelligence journal last month that “Reward is enough” to reach general AI. The paper was first reported by VentureBeat last week.

In the paper, the researchers claim that if you keep “rewarding” an algorithm each time it does something you want it to, which is the essence of reinforcement learning, then it will eventually start to show signs of general intelligence.

“Reward is enough to drive behavior that exhibits abilities studied in natural and artificial intelligence, including knowledge, learning, perception, social intelligence, language, generalization and imitation,” the authors write.

“We suggest that agents that learn through trial and error experience to maximize reward could learn behavior that exhibits most if not all of these abilities, and therefore that powerful reinforcement learning agents could constitute a solution to artificial general intelligence.”

Not everyone is convinced, however.

Samim Winiger, an AI researcher in Berlin, told CNBC that DeepMind’s “reward is enough” view is a “somewhat fringe philosophical position, misleadingly presented as hard science.”

He said the path to general AI is complex and that the scientific community is aware that there are countless challenges and known unknowns that “rightfully instill a sense of humility” in most researchers in the field and prevent them from making “grandiose, totalitarian statements” such as “RL is the final answer, all you need is reward.”

DeepMind told CNBC that while reinforcement learning has been behind some of its most well-known research breakthroughs, the AI technique accounts for only a fraction of the overall research it carries out. The company said it thinks it’s important to understand things at a more fundamental level, which is why it pursues other areas such as “symbolic AI” and “population-based training.”

“In somewhat typical DeepMind fashion, they chose to make bold statements that grabs attention at all costs, over a more nuanced approach,” said Winiger. “This is more akin to politics than science.”

Stephen Merity, an independent AI researcher, told CNBC that there’s “a difference between theory and practice.” He also noted that “a stack of dynamite is likely enough to get one to the moon, but it’s not really practical.”

Ultimately, there’s no proof either way to say whether reinforcement learning will ever lead to AGI.

Rodolfo Rosini, a tech investor and entrepreneur with a focus on AI, told CNBC: “The truth is nobody knows and that DeepMind’s main product continues to be PR and not technical innovation or products.”

Entrepreneur William Tunstall-Pedoe, who sold his Siri-like app Evi to Amazon, told CNBC that even if the researchers are correct “that doesn’t mean we will get there soon, nor does it mean that there isn’t a better, faster way to get there.”

DeepMind’s “Reward is enough” paper was co-authored by DeepMind heavyweights Richard Sutton and David Silver, who met DeepMind CEO Demis Hassabis at the University of Cambridge in the 1990s.

“The key problem with the thesis put forth by ‘Reward is enough’ is not that it is wrong, but rather that it cannot be wrong, and thus fails to satisfy Karl Popper’s famous criterion that all scientific hypotheses be falsifiable,” said a senior AI researcher at a large U.S. tech firm, who wished to remain anonymous due to the sensitive nature of the discussion.

“Because Silver et al. are speaking in generalities, and the notion of reward is suitably underspecified, you can always either cherry pick cases where the hypothesis is satisfied, or the notion of reward can be shifted such that it is satisfied,” the source added.

“As such, the unfortunate verdict here is not that these prominent members of our research community have erred in any way, but rather that what is written is trivial. What is learned from this paper, in the end? In the absence of practical, actionable consequences from recognizing the unalienable truth of this hypothesis, was this paper enough?”

What is AGI?

While AGI is often referred to as the holy grail of the AI community, there’s no consensus on what AGI actually is. One definition is it’s the ability of an intelligent agent to understand or learn any intellectual task that a human being can.

But not everyone agrees with that and some question whether AGI will ever exist. Others are terrified about its potential impacts and whether AGI would build its own, even more powerful, forms of AI, or so-called superintelligences.

Ian Hogarth, an entrepreneur turned angel investor, told CNBC that he hopes reinforcement learning isn’t enough to reach AGI. “The more that existing techniques can scale up to reach AGI, the less time we have to prepare AI safety efforts and the lower the chance that things go well for our species,” he said.

Winiger argues that we’re no closer to AGI today than we were several decades ago. “The only thing that has fundamentally changed since the 1950/60s, is that science-fiction is now a valid tool for giant corporations to confuse and mislead the public, journalists and shareholders,” he said.

Fueled with hundreds of millions of dollars from Alphabet every year, DeepMind is competing with the likes of Facebook and OpenAI to hire the brightest people in the field as it looks to develop AGI. “This invention could help society find answers to some of the world’s most pressing and fundamental scientific challenges,” DeepMind writes on its website.

DeepMind COO Lila Ibrahim said on Monday that trying to “figure out how to operationalize the vision” has been the biggest challenge since she joined the company in April 2018.

Continue Reading

Technology

Hinge founder leaves CEO role to launch AI-powered dating startup

Published

on

By

Hinge founder leaves CEO role to launch AI-powered dating startup

Justin McLeod speaks during the Fast Company Innovation Festival 2025 on Sept. 18, 2025 in New York City.

Eugene Gologursky | Getty Images

Hinge founder Justin McLeod is stepping down as CEO of the dating app to launch a dating service powered by artificial intelligence.

McLeod will be replaced by Jackie Jantos, the dating app’s president and chief marketing officer, Hinge parent company Match Group announced on Tuesday.

“The company’s momentum, including being on track to reach $1 billion in revenue by 2027, gives me full confidence in where Hinge is headed,” said McLeod in a statement. He created the dating app in 2011.

McLeod will remain as an advisor to Hinge through March. Overtone, his new venture, will use AI and voice tools to “help people connect in a more thoughtful and personal way,” according to the announcement.

Along with a dedicated team, McLeod spent much of this year developing the startup with support from Match Group, which said it plans to lead Overtone’s initial funding round in early 2026.

Match Group, which also owns Tinder and various other dating apps, will hold a significant ownership position in Overtone. Match Group CEO Spencer Rascoff will join Overtone’s board.

“We’re proud to have incubated Overtone within Hinge and to now lead its funding round as he builds his next venture,” Rascoff said in a statement.

WATCH: Software could start benefitting from AI in 2026, says Intelligent Alpha CEO Doug Clinton

Software could start benefitting from AI in 2026, says Intelligent Alpha CEO Doug Clinton

Continue Reading

Technology

Oracle’s AI-fueled debt load has investors on edge ahead of quarterly earnings

Published

on

By

Oracle's AI-fueled debt load has investors on edge ahead of quarterly earnings

Oracle CEOs Clay Magouyrk and Mike Sicilia sit down with CNBC’s David Faber on Oct. 13, 2025.

CNBC

It’s been a rollercoaster year for Oracle investors, as they try to assess the strength of the software giant’s position in the artificial intelligence boom.

The stock is up more than 30% for the year even after a 23% plunge in October, which was its worst month since 2001. It’s recovered a bit in November, climbing almost 10% for the month as of Tuesday.

Heading into the company’s fiscal second-quarter earnings report on Wednesday, pressure is building on management — and newly installed CEOs Clay Magouyrk and Mike Sicilia — to show that Oracle can continue to finance the company’s aggressive infrastructure plans while simultaneously convincing Wall Street that the AI-fueled hypergrowth story remains intact.

In recent months, Oracle has emerged as a more central player in AI, largely due to a $300 billion deal with OpenAI, which came to light in September, an agreement that involves the AI startup buying computing power over about five years, starting in 2027.

Funding Oracle’s compute buildout is going to require mounds of debt. In late September, Oracle raised $18 billion in a jumbo bond sale, one of the largest debt issuances on record in the tech industry, and the company is now the biggest issuer of investment grade debt among non-financial firms, according to Citi.

“There is something inherently uncomfortable as a credit investor about the transformation of the sort we’re facing that is going to require an enormous amount of capital,” Daniel Sorid, head of U.S. investment grade credit strategy at Citi, said on a video call to investors on Friday, a replay of which was provided to reporters.

Oracle's new capacity lags behind competitors, says Rothschild's Haissl

Oracle has secured billions of dollars of construction loans through a consortium of banks tied to data centers in New Mexico and Wisconsin. Citi analyst Tyler Radke estimates Oracle will raise roughly $20 billion to $30 billion in debt every year for the next three years.

As of August, the company’s combined short-term and long-term debt, which includes lease obligations, sat at $111.6 billion, up from $84.5 billion a year earlier, according to FactSet, while cash and equivalents slipped over that stretch to $10.45 billion from $10.6 billion.

As Oracle aims to build out sufficient capacity to meet the rising demand its seeing from customers like OpenAI, the street is questioning whether company will tap sources other than the debt market.

“Oracle will be looking at all options out there — off-balance sheet facilities, raising debt, issuing equity or perhaps exploring interest from a foreign investor, i.e. a sovereign wealth fund,” said Rishi Jaluria, a software analyst at RBC Capital Markets, in an interview. Jaluria recommends holding the stock.

A credit investor who spoke to CNBC highlighted Meta’s $27 billion deal with Blue Owl Capital, a joint venture between the two entities, as one type of financing arrangement being used for AI data center development.

The market is also debating whether Oracle can use vendor financing options to reduce the amount of upfront capital required to stand up data centers, including securing favorable financing terms with suppliers like Nvidia, a credit investor told CNBC. However in that scenario, Nvidia’s chips would be used as collateral, raisings concerns around GPU depreciation.

An Oracle spokesperson declined to comment.

Growing skepticism

The discomfort that Sorid referenced has driven Oracle’s 5-year credit default swaps to new multi-year highs. Credit default swaps are like insurance for investors, with buyers paying for protection in case the borrower can’t repay its debt. Bond investors told CNBC that they’ve become a popular way to hedge the risk tied to the AI trade.

Credit analysts at Barclays and Morgan Stanley are recommending clients buy Oracle’s 5-year CDS. Andrew Keches, an analyst at Barclays, told analysts in a note last month that he didn’t see an avenue for Oracle’s credit trajectory to improve. And in late November, Morgan Stanley analysts said Oracle’s CDS had attracted not just typical credit investors but “tourists” who have less experience with this type of financial instrument.

Spools of electrical wires outside a series of assembly tents during a media tour of the Stargate AI data center in Abilene, Texas, US, on Tuesday, Sept. 23, 2025. Stargate is a collaboration of OpenAI, Oracle and SoftBank, with promotional support from President Donald Trump, to build data centers and other infrastructure for artificial intelligence throughout the US.

Kyle Grillot | Bloomberg | Getty Images

Oracle’s revenue growth and backlog of business will be closely monitored as investors try to gauge whether the company’s spending plans are justified. Analysts expect to see revenue growth in the latest quarter of 15% to $16.2 billion, according to StreetAccount.

Remaining performance obligations, a measure of contracted revenue that hasn’t yet been recognized, are expected to surpass $500 billion, StreetAccount says, which would mark a more than fivefold increase from a year earlier. Oracle’s disclosure in September that RPOs jumped 359% to $455 billion sent the company’s stock up 36%, its best single-day performance since 1992.

Since then, the stock has wiped out all of those gains and then some.

Gil Luria, an analyst at D.A. Davidson, said that beyond infrastructure, he’ll be closely watching Oracle’s core database business, which is a source of much higher margins. That will help determine how much flexibility the company has in going to the capital markets, he said.

“Oracle can handle the debt load,” said Luria, who recommends holding the stock. “But they need more cash flow to raise more capital from here.”

WATCH: Oracle has to talk execution, ‘not just promises’

Oracle has to talk execution, 'not just promises', says DCLA's Sarat Sethi

Continue Reading

Technology

Teachers’ union AFT slams crypto market bill, warns of ‘profound risks’ for America’s retirement plans

Published

on

By

Teachers' union AFT slams crypto market bill, warns of 'profound risks' for America's retirement plans

Sen. Gillibrand says 'nothing is holding up' progress on crypto market regulation: CNBC Crypto World

The American Federation of Teachers, the powerful labor union that represents 1.8 million members, is urging the Senate Banking Committee to reconsider its crypto market structure bill, the Responsible Financial Innovation Act, calling the proposed legislation “as irresponsible as it is reckless” in a letter exclusively obtained by CNBC.

In the letter that AFT president Randi Weingarten sent to Senate Banking Committee Chairman Tim Scott (R-SC) and Ranking Member Elizabeth Warren (D-Mass.), she wrote the union opposes the bill based on the “profound risks to the pensions of working families and the overall stability of the economy.”

“The legislation on crypto we have seen weighed by the committee over the last few months gives us deep concern,” Weingarten added.

The AFT is concerned that in passing crypto legislation, the government will open the floodgates to widespread fraud and unethical practices across retirement plans including AFT pensions.

“This legislation pretends that crypto assets are stable and mainstream, and they are not. Rather than just being silent on crypto, this bill strips the few safeguards that exist for crypto and erodes many protections for traditional securities. If passed, it will undercut the safety of many assets and cause problems across retirement investments,” Weingarten wrote.

A specific issue the AFT cited with the proposed legislation it allowing non-crypto companies to put their stock on the blockchain and evade existing securities regulatory framework. Wall Street has become interested in the idea of “tokenization” of all financial assets, with Larry Fink, CEO of BlackRock, the largest asset manager in the world, a leader evangelist for the concept.

“This loophole and the erosion of traditional securities law will have disastrous consequences: Pensions and 401(k) plans will end up having unsafe assets even if they were invested in traditional securities,” Weingarten wrote.

She argued that the legislation being considered by the committee also does little to curb fraud, illegal activity and corruption that continues to be prevalent in crypto markets. Weingarten called the legislation “irresponsible” and “reckless.”

“We believe that if enacted, this bill has the potential to lay the groundwork for the next financial crisis,” she wrote.

NEW YORK, NEW YORK – AUGUST 28: Randi Weingarten, president of the American Federation of Teachers (AFT), speaks during the March on Wall Street on August 28, 2025 in New York City.

Michael M. Santiago | Getty Images News | Getty Images

The AFL-CIO, the nation’s largest labor union, stated its opposition to the Senate Banking Committee over a draft of the crypto bill in October.

CNBC also confirmed that on Thursday, the CEOs of Bank of America, Citi and Wells Fargo, will be meeting with lawmakers to discuss the crypto market structure proposals.

The currently proposed legislation, which builds on a bill that passed the House of Representatives over the summer, is co-sponsored by key crypto backer Senator Cynthia Lummis (R-Wyoming) and Senator Bernie Moreno (R-Ohio), alongside Chairman Scott. It aims to create structure for regulating digital assets, but also raises questions about tokenized securities that are not specifically cryptocurrencies.

Tokenization has been a key concern as the bill has gained momentum on Capitol Hill, and a hurdle to getting the support from Democrats that will be needed for passage. Previous CNBC reporting indicates that the Senate backers will need to attract votes from at least seven Democrats for the legislation to pass. At last week’s CNBC CFO Council Summit in Washington, D.C., Senator Mark Warner (D-Va.) told attendees, “I’m in crypto hell at this moment trying to get the market structure bill done.”

Warner is among a group of Democratic senators who met on Monday to review the Senate Banking draft and consider counter-offers, according to Politico.

Many Democrats, including Warren, have also been concerned about the balance of crypto regulatory oversight between the CFTC and the Securities and Exchange Commission. States, meanwhile, worry that their laws may be preempted by a new federal law, and the states left powerless to protect residents from fraud, a concern outlined by Massachusetts’ Secretary of State William Galvin in a letter to Senate Banking, writing that the “sweeping provisions that will exclude significant portions of the financial industry from state oversight. This is a recipe for disaster for millions of savers.”

Progress on the Senate’s version of a crypto market structure bill was stalled for weeks due to the longest government shutdown in U.S. history. Speaking on Tuesday morning at The Blockchain Association Policy Summit in Washington, D.C., Senator Lummis provided some insight into when the Senate’s version of a crypto market structure bill could be expected. She said her goal is to share a draft by the end of the week, then let the crypto industry as well as Republicans and Democrats vet it and proceed to markup next week.

CFTC announces listed spot crypto trading on U.S. regulated exchanges: CNBC Crypto World

Continue Reading

Trending