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

Sam Bankman-Fried tried to influence witness through Signal, DOJ alleges

Published

on

By

Sam Bankman-Fried tried to influence witness through Signal, DOJ alleges

Former FTX chief executive Sam Bankman-Fried (C) arrives to enter a plea before US District Judge Lewis Kaplan in the Manhattan federal court, New York, January 3, 2023. 

Ed Jones | AFP | Getty Images

Federal prosecutors are attempting to bar indicted FTX co-founder Sam Bankman-Fried from using encrypted messaging software, citing efforts that may “constitute witness tampering,” according to a letter filed in Manhattan federal court Friday.

Bankman-Fried reached out to the “current General Counsel of FTX US who may be a witness at trial,” prosecutors said. Ryne Miller, who was not identified by name in the government filing, is the current counsel for FTX US, and a former partner at Kirkland & Ellis.

The government claims that Bankman-Fried wrote to Miller via Signal, an encrypted messaging app, on Jan. 15, days after bankruptcy officials at crypto exchange disclosed the recovery of more than $5 billion in FTX assets.

“I would really love to reconnect and see if there’s a way for us to have a constructive relationship, use each other as resources when possible, or at least vet things with each other,” Bankman-Fried allegedly told Miller.

Bankman-Fried has also been in contact with “other current and former FTX employees,” the filing said. Federal prosecutors allege that Bankman-Fried’s request suggests an effort to influence the witness’s testimony, and that Bankman-Fried’s effort to improve his relationship with Miller “may itself constitute witness tampering.”

Both Miller and a representative for Bankman-Fried declined to comment.

In restricting Bankman-Fried’s access to Signal and other encrypted messaging platforms, the government cites a need to “prevent obstruction of justice.” Federal prosecutors claim that Bankman-Fried directed Alameda and FTX through Slack and Signal, and ordered his employees set communications to “autodelete after 30 days or less.”

Citing previously undisclosed testimony from ex-Alameda CEO Caroline Ellison, the government claimed that Bankman-Fried indicated “many legal cases turn on documentation and it is more difficult to build a legal case if information is not written down or preserved.” Ellison pled guilty to multiple charges of fraud and has been cooperating with the U.S. Attorney’s efforts to build a case against Bankman-Fried.

Bankman-Fried pled not guilty to eight charges in connection with the collapse of his multibillion-dollar crypto empire, FTX. He is due in federal court in October, after being released on $250 million bond.

Continue Reading

Technology

Amazon to start charging delivery fees on Fresh grocery orders under $150

Published

on

By

Amazon to start charging delivery fees on Fresh grocery orders under 0

Brendan McDermid | Reuters

Amazon will start charging delivery fees for Fresh grocery orders that are less than $150, in a move it said will help keep prices low on its services.

Beginning Feb. 28, Prime members who want home delivery from Amazon Fresh will incur a $9.95 delivery fee for orders under $50, while orders between $50 and $100 will include a $6.95 delivery fee, and orders between $100 and $150 will carry a $3.95 delivery fee, the company said in a note to customers viewed by CNBC. Only Prime members can use the delivery service, although anybody can shop at an Amazon Fresh grocery store.

related investing news

Here are Thursday's biggest analyst calls: Tesla, Apple, Roku, Peloton, Plug Power, Target & more

CNBC Pro

Amazon previously guaranteed members of its $139-a-year Prime service free delivery on Fresh orders over $35.

“This service fee will help keep prices low in our online and physical grocery stores as we better cover grocery delivery costs and continue to enable offering a consistent, fast, and high-quality delivery experience,” the notice stated.

The move comes as Amazon CEO Andy Jassy has embarked on a wide-ranging review of the company’s expenses amid slowing sales and a worsening economic outlook. Amazon has eyed laying off 18,000 employees, frozen hiring in its corporate workforce, and paused or canceled some projects such as a sidewalk robot and a telehealth service.

Amazon has previously recalibrated its approach to online grocery deliveries, a business that is notoriously challenging from a cost and efficiency perspective. In 2021, Amazon added a $10 service fee for Whole Foods delivery orders to Prime members, after previously offering them for no extra charge.

WATCH: How Whole Foods has changed in the five years since Amazon took over

How Whole Foods has changed in the five years since Amazon took over

Continue Reading

Technology

Tesla just had its best week since May 2013

Published

on

By

Tesla just had its best week since May 2013

Tesla CEO Elon Musk smiles as he addresses guests at the Offshore Northern Seas 2022 (ONS) meeting in Stavanger, Norway on August 29, 2022.

Carina Johansen | AFP | Getty Images

Tesla shares surged 33% this week, marking their best weekly performance since May 2013 and second best on record.

The stock rose 11% on Friday to close at $177.88. The rebound followed a six-month period in which Tesla shares had declined more than 40%. The stock’s 65% plunge in 2022 was its worst in Tesla’s 12-plus years as a public company.

Tesla’s rally this week was aided by an upbeat fourth-quarter earnings report. During the call with shareholders and analysts, CEO Elon Musk said the company was on target to potentially produce 2 million vehicles in 2023, and he suggested demand would support sales of those cars as well.

Official guidance called for production of 1.8 million vehicles this year. The company has not revised its longstanding target for 50% compound annual growth rate over a multi-year horizon.

Tesla’s five day performance charted against Rivian and Ford Motor Company.

Tesla beat on both the top and the bottom lines, recording total revenue of $24.32 billion, including $324 million of deferred revenue related to Tesla’s driver assistance systems. The company cut prices for its cars dramatically in December and January, leading to concern about demand and a buildup of inventory.

Analyst reaction to Tesla’s numbers was mixed.

“For bulls, the growth story is alive and well,” Bernstein’s Toni Sacconaghi, who has an underperform rating on the stock, wrote in a note on Thursday. “For bears, the numbers don’t lie.”

In early January, Tesla reported fourth-quarter vehicle deliveries and production that fell shy of expectations.

Tesla’s stock jump came amid a broader market rally. The S&P 500 was up 2.2% for the week and the Nasdaq gained 4.3%.

Other U.S.-based electric vehicle makers saw their shares climb higher. Rivian rose 22% during the week, while shares in legacy automakers Ford and General Motors each gained more than 7%.

Rival electric car manufacturer Lucid spiked on Friday as well, rising 43% on reports of rumors that Saudi Arabia’s sovereign wealth fund, the Public Investment Fund, intended to take the company private.

Some of Tesla’s underperformance last year was attributed to Musk’s shift of focus to Twitter, which he acquired for $44 billion in October. Under Musk’s leadership, Twitter has experienced mass layoffs and fleeing advertisers, gutting morale.

Tesla remains the second most-shorted stock in U.S. markets, behind only Apple, meaning that a large numbers of investors are betting on a decline. Over 94 million of the automaker’s shares are shorted, according to data from S3 Partners.

Despite the rally, active short selling continues, S3 managing director Ihor Dusaniwsky told CNBC. Short sellers view Tesla’s appreciation as having created “an overheated and overbought stock that is due for at least a short-term reversal,” he said. In the last week, S3 Partners said it’s seen a 3.9% increase in total shares shorted, while investors shorting the stock lost $4.3 billion over that stretch.

WATCH: Tesla still in league of its own

Why Tesla has always been an anomaly in the automotive industry

Continue Reading

Trending