Computer scientists are questioning whether Alphabet’s DeepMind will ever make A.I. more human-like

Published

4 years ago

June 18, 2021

admin

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.

Technology

SoFi’s stock drops on $1.5 billion share sale announcement

Published

4 hours ago

December 4, 2025

admin

SoFi's stock drops on .5 billion share sale announcement

Anthony Noto, CEO of SoFi, speaking with CNBC at the annual Allen & Co. Media and Technology Conference in Sun Valley, Idaho on July 10th, 2025.

David A. Grogan | CNBC

SoFi shares fell almost 6% in extended trading Thursday after the fintech company announced a $1.5 billion stock offering.

The company, which provides online loans and other banking services, said in a press release that it will use the proceeds for “general corporate purposes, including but not limited to enhancing capital position, increasing optionality and enabling further efficiency of capital management, and funding incremental growth and business opportunities.”

The announced offering comes after SoFi’s market cap almost doubled so far in 2025. The stock price is up more than sixfold since the end of 2022.

A company’s share price often drops on a planned share sale as the offering dilutes the value of existing holders’ stakes.

In its third-quarter earnings release in late October, SoFi reported revenue growth of 38% from a year earlier to $961.6 million, while net income more than doubled to $139.4 million. The company reported cash and equivalents of $3.25 billion.

WATCH: SoFi CEO on launch of crypto trading

Technology

Apple announces departure of general counsel and policy chief

Published

6 hours ago

December 4, 2025

admin

Apple announces departure of general counsel and policy chief

Lisa Jackson, senior vice president of environment, policy and social initiatives at Apple Inc., speaks during the TechCrunch Disrupt 2017 in San Francisco, California, U.S., on Tuesday, Sept. 19, 2017.

David Paul Morris | Bloomberg | Getty Images

Apple’s general counsel, Kate Adams, and its vice president for environment, policy, and social initiatives, Lisa Jackson, are retiring from the company, the iPhone maker announced on Thursday.

Jennifer Newstead, Meta’s chief legal officer, will become Apple’s new general counsel in March, and Jackson’s government affairs staff will report to her starting late next year, Apple said.

The two executives, who both reported to Apple CEO Tim Cook, are the latest members of senior leadership to exit the company. In recent weeks, Apple’s head software designer said he was leaving to join Meta, while Apple said its AI chief was retiring, along with its chief operating officer.

Adams joined Apple from Honeywell and became general counsel in 2017, and oversaw legal matters including litigation, global security, and the company’s privacy initiatives. Under Adams, Apple grappled with rising antitrust scrutiny and regulation around the world, including major lawsuits in the U.S. over the iPhone App Store’s restrictions and fees.

Jackson joined Apple in 2013, and led the company’s diversity programs as well as much of its policy work in Washington, D.C. Before that, she spent four years as administrator of the U.S. Environmental Protection Agency, a position she was appointed to by President Barack Obama.

With her emphasis in areas like social justice and renewable energies, Jackson’s job lost relevance during the second Trump administration, which has publicly denounced diversity, equity and inclusion programs and slammed efforts to combat climate change.

Apple has faced increased tariffs from the Trump administration, and Cook has met with President Donald Trump several times to tout the company’s American manufacturing plans as part of an effort to influence policy.

Jackson was instrumental in Apple’s launch of its Racial Equity and Justice Initiative following the 2020 murder of George Floyd. She then helped expand the company’s equity and justice efforts to other countries, including the U.K., Mexico and New Zealand, according to a report published in 2023.

“At Apple, we pledge that our resolve will not fade,” Jackson wrote in a section of that report. “We won’t delay action. We will work, each and every day, on the urgent task of advancing equity.”

Jackson also worked on Apple’s environmental image. Her job “focused on reducing greenhouse gases, protecting air and water quality, preventing exposure to toxic contamination, and expanding outreach to communities on environmental issues,” according to her bio on the company’s website. She discussed Apple’s plans to become carbon neutral at iPhone launch events.

Jackson also accompanied Cook to several official functions in Washington, including state dinners.

Apple CEO Tim Cook and Apple Vice President Lisa Jackson arrive at the White House for a state dinner on April 10, 2024 in Washington, DC.

Tasos Katopodis | Getty Images

Newstead, who will become Apple’s top lawyer, has overseen Meta’s legal and regulatory matters pertaining to its family of apps like Facebook, Instagram, WhatsApp since 2019. A Meta spokesperson said Newstead will be staying through the end of the year and that the company is actively searching for her replacement.

Prior to Meta, Newstead served as a Trump-appointed legal advisor at the State Department during the president’s first administration in 2019.

Before that, she was a partner at Davis Polk & Wardwell and a general counsel of the White House Office of Management and Budget, among other roles in the U.S. government.

CNBC’s Jonathan Vanian contributed to this story.

WATCH: Apple AI chief steps down

Apple AI chief steps down: Here's what to know

Technology

Microsoft will raise prices of commercial Office subscriptions in July

Published

9 hours ago

December 4, 2025

admin

Microsoft will raise prices of commercial Office subscriptions in July

A general view of the Microsoft office building is seen in Cologne, Germany, on November 18, 2025.

Nurphoto | Nurphoto | Getty Images

Microsoft said Thursday that it will increase the prices of Office productivity software subscriptions for commercial and government clients on July 1.

The company’s Office applications, which include Word, Excel, PowerPoint and Outlook, have been facing increased competition in recent years from Google.

“We are continuously investing and innovating our platform for the future,” Nicole Herskowitz, corporate vice president for Microsoft 365 and Copilot, wrote in a blog post. “In the last year, we released more than 1,100 features across Microsoft 365, Security, Copilot, and SharePoint.” The new features have added value to the suites, she wrote.

Price hikes for commercial Office subscriptions have been infrequent. In 2022, Microsoft raised prices of its productivity bundles for the first time since launching the original Office 365 subscriptions in 2011. Microsoft changed the name of Office 365 to Microsoft 365 in 2020. In January, Microsoft announced a price hike for consumer Office bundles.

Microsoft offers Office 365 subscriptions for commercial use that include access to its productivity applications, along with higher-priced Microsoft 365 subscriptions that also include Windows operating system updates.

Here’s a breakdown of the commercial price changes:

For small and medium-sized businesses, Microsoft 365 Business Basic will cost $7 per person per month, up from $6.
Microsoft 365 Business Standard will be available for $14, up from $12.50.
Microsoft 365 Business Premium will continue to cost $22.
The entry-level Office 365 E1 offering for enterprises will still be sold for $10.
Office 365 E3 will jump 13% to $26 from $23.
The Microsoft 365 E3 package including Windows for enterprises will rise 8% to $39 from $36.
The full-featured Microsoft 365 E5 will increase to $60 from $57.
For front-line workers such as cashiers, Microsoft 365 F1 subscriptions will cost $3, up from $2.25.
Microsoft 365 F3 will be available for $10, up from $8.

The U.S. Defense Department and other government clients will face similar percentage price increases.

The various subscriptions all exclude access to the $30 Microsoft 365 Copilot add-on that draws on generative artificial intelligence models. Some companies have started widely rolling out Copilot, while others have held off on expanding their deployments, CNBC reported last week.

In many cases, organizations receive discounts off of list prices, but Microsoft has cut back on direct volume deals for some types of customers.

Almost 43% of Microsoft’s $77.7 billion in fiscal first-quarter revenue came from its Productivity and Businesses Processes segment, which includes Office. In October, the company said revenue from Microsoft 365 commercial cloud services jumped 17%, while seats increased 6%, mainly from products targeting small and medium-sized businesses and front-line workers.

WATCH: Microsoft sees ‘huge’ challenge and great opportunity as global economy enters a new phase, president says