Science & Technology
Google AI chief says there’s a 50% chance we’ll hit AGI in just 5 years
More than a decade ago, the co-founder of Google’s DeepMind artificial intelligence lab predicted that by 2028, AI will have a half-and-half shot of being about as smart as humans — and now, he’s holding firm on that forecast.
In an interview with tech podcaster Dwarkesh Patel, DeepMind co-founder Shane Legg said that he still thinks that researchers have a 50-50 chance of achieving artificial general intelligence (AGI), a stance he publicly announced at the very end of 2011 on his blog, Futurism reported.
It’s a notable prediction considering the exponentially growing interest in the space. OpenAI CEO Sam Altman has long advocated for an AGI, a hypothetical agent that is capable of accomplishing intellectual tasks as well as a human, that can be of benefit to all. But whether we’ll ever be able to get to that point — let alone agree on one definition of AGI — remains to be seen.
Legg apparently began looking towards his 2028 goalpost all the way back in 2001 after reading “The Age of Spiritual Machines,” the groundbreaking 1999 book by fellow Google AI luminary Ray Kurzweil that predicts a future of superhuman AIs.
“There were two really important points in his book that I came to believe as true,” he explained. “One is that computational power would grow exponentially for at least a few decades. And that the quantity of data in the world would grow exponentially for a few decades.”
Paired with an understanding of the trends of the era, such as the deep learning method of teaching algorithms to “think” and process data the way human brains do, Legg wrote back at the start of the last decade that in the coming ones, AGI could well be achieved — so long as “nothing crazy happens like a nuclear war.”
Today, the DeepMind co-founder said that there are caveats to his prediction that the AGI era will be upon us by the end of this decade.
The first, broadly, is that definitions of AGI are reliant on definitions of human intelligence — and that kind of thing is difficult to test precisely because the way we think is complicated.
“You’ll never have a complete set of everything that people can do,” Legg said — things like developing episodic memory, or the ability to recall complete “episodes” that happened in the past, or even understanding streaming video. But if researchers could assemble a battery of tests for human intelligence and an AI model were to perform well enough against them, he continued, then “you have an AGI.”
When Patel asked if there could be a single simple test to see whether an AI system had reached general intelligence, such as beating Minecraft, Legg pushed back.
“There is no one thing that would do it, because I think that’s the nature of it,” the AGI expert said. “It’s about general intelligence. So I’d have to make sure [an AI system] could do lots and lots of different things and it didn’t have a gap.”
The second biggest caveat, Legg added, was the ability to scale AI training models way, way up — a worthy point given how much energy AI companies are already using to churn out large language models like OpenAI’s GPT-4.
“There’s a lot of incentive to make a more scalable algorithm to harness all this computing data,” Legg explained. “So I thought it would be very likely that we’ll start to discover scalable algorithms to do this.”
Asked where he thought we stand today on the path to AGI, Legg said that he thinks computational power is where it needs to be to make it happen, and the “first unlocking step” would be to “start training models now with the scale of the data that is beyond what a human can experience in a lifetime” — a feat he believes the AI industry is ready to achieve.
All that said, Legg reiterated his personal stance that he only believes there’s a 50 percent chance researchers will achieve AGI before the end of this decade, and Futurism has reached out to DeepMind to see if the Google subsidiary has anything to add to that prognosis.
“I think it’s entirely plausible,” he said, “but I’m not going to be surprised if it doesn’t happen by then.”
Science & Technology
Musk’s Starlink faces high-profile security test in Iran crackdown
Starlink, which is harder for Iran to tamper with than cable and cellphone tower networks, has become crucial for documenting events on the ground.
Iran’s crackdown on dissidents is shaping up as one of the toughest security tests yet for Elon Musk’s Starlink, which has served as a lifeline against state-imposed internet blackouts since its deployment during the war in Ukraine, Reuters reported.
SpaceX, which owns Starlink, made the satellite service free for Iranians this week, placing Musk’s space company at the center of another geopolitical hot spot and pitting a team of U.S.-based engineers against a regional power armed with satellite jammers and signal-spoofing tactics, according to activists, analysts and researchers.
How SpaceX withstands Iranian attacks on its most lucrative line of business is expected to be closely watched by U.S. military forces and intelligence agencies that use Starlink and its military-grade variant Starshield, as well as China, whose own nascent satellite internet constellations are set to rival Starlink in the coming years. With SpaceX weighing a public listing this year, the situation in Iran also represents a high-profile showcase for Starlink to investors.
“We’re in this weird early part of the history of space-delivered communications where SpaceX is the only true provider at this scale,” said John Plumb, the former Pentagon space policy chief under President Joe Biden.
“And these repressive regimes think they can still turn off communications, but I think the day is coming where that’s just not possible,” he said.
Victoria Samson, chief director of space security and stability at the think tank Secure World Foundation, said Russia, which has deployed an array of technologies to counter Starlink in Ukraine, might be keen to examine the effectiveness of Iran’s Starlink interference.
“I think a lot of actors are watching how Starlink fares here,” she said.
Thousands of people protesting Iran’s clerical rule are reported to have been killed in the past week, while Tehran’s order to restrict communications makes it difficult to discern the full extent of its violent crackdown on dissent.
Starlink, which is harder for Iran to tamper with than cable and cellphone tower networks, has become crucial for documenting events on the ground, read the report.
Raha Bahreini, an Iran researcher at Amnesty International, said they had verified dozens of videos from Iran, including footage of protesters killed or injured by Iranian forces, and believe that almost all of them came from people who had access to Starlink. She added, however, that the ongoing communications restrictions have impeded human rights organizations’ communications with people in Iran in efforts to assess the scale of the violence.
Starlink is banned in Iran, yet tens of thousands of terminals may have been smuggled into the country, although it remains unclear how many are in use, according to Holistic Resilience, a U.S. nonprofit that has helped deliver Starlink terminals to Iranians and says it is working with SpaceX to monitor what it describes as Iranian attempts to jam the system.
Consumer Starlink terminals are rectangular antenna dishes that come in two sizes – one roughly the size of a pizza box and a smaller “mobile” one the size of a laptop.
SpaceX did not respond to requests for comment.
The Iranian mission to the United Nations in New York declined to comment on Thursday in response to Reuters’ questions.
Iranian Foreign Minister Abbas Araqchi, speaking to Al Jazeera TV on Monday, said the internet had been cut off “after we confronted terrorist operations and realized orders were coming from outside the country.”
Starlink, the first massive internet-from-space constellation of its kind, has emerged as a crucial tool for communications in wartime and remote areas. The network, which drove SpaceX’s $15 billion revenue in 2024, has expanded the geopolitical power of Musk, who in 2022 asserted control over how and where it was being used by Ukrainian troops fighting back Russian forces.
Roughly 10,000 low-orbiting Starlink satellites zipping above user terminals at an orbital velocity of some 17,000 miles per hour (27,360 kph) make its signals much harder to locate and disrupt than traditional satellite systems designed with a larger, single satellite fixed over a given territory.
Iran is likely using satellite jammers to disrupt the Starlink signals, according to Holistic Resilience and other specialists. Iran also appears to be engaging in so-called spoofing, or broadcasting fake GPS signals to confuse and disable Starlink terminals, according to Nariman Gharib, an Iranian opposition activist and independent cyber espionage investigator based in Britain, Reuters reported.
The GPS spoofing wreaks havoc on a Starlink terminal’s connection and slows internet speeds, said Gharib, who analyzed data from a terminal inside Iran.
“You might be able to send text messages, but forget about video calls,” he said.
Though Starlink is not licensed to operate in Iran, Musk has repeatedly confirmed its presence on his social media platform X, spurring a yearslong effort by the Iranian government to counter the service. Amid protests over the death of Mahsa Amini in December 2022, Musk posted that nearly 100 Starlink terminals were active in the country.
Following the 12-day war between Iran and Israel in June, Iran’s parliament passed a law banning the use of Starlink, introducing severe penalties for those who use or distribute the unlicensed technology, according to Iranian state media.
Iran has also pursued diplomatic channels, urging a panel at the United Nations’ International Telecommunication Union last year to force the United States and Norway — where Starlink is internationally registered — to block the service.
At a July meeting, Iran told the board that Starlink’s use in the country is illegal and said an “invading country” had deployed its terminals on drones during a recent attack.
Iran told the board in November that it was struggling to locate and disable the terminals itself.
Science & Technology
Indian rocket launch loses control after liftoff in fresh blow to ISRO
An Indian rocket carrying 16 loads of equipment and experiments including an earth surveillance satellite went off track after liftoff on Monday in a fresh setback to the workhorse launch vehicle of the Indian Space Research Organisation.
It was a second disappointment for the Polar Satellite Launch Vehicle in about eight months, denting its reputation for reliability, with a more than 90% success rate over about 60 past missions, Reuters reported.
The PSLV-C62 lifted off from the Satish Dhawan Space Centre on the island of Sriharikota at 10:18 a.m. (04488 GMT) carrying the EOS-N1 observation satellite and 15 other payloads developed by startups and academic institutions in India and abroad.
The ISRO’s mission control said the rocket performed normally for most of the flight before an unexpected disturbance and deviation from its path.
“The PSLV-C62 mission encountered an anomaly during the end of the PS3 stage. A detailed analysis has been initiated,” ISRO said in a statement, without giving further details on what had gone wrong or where the rocket ended up.
The PSLV has been central to India’s space programme, having launched missions such as Chandrayaan-1 and the Aditya-L1 solar observatory. It also underpins India’s push to open space manufacturing to private industry.
Science & Technology
EU considers making WhatsApp more responsible for tackling harmful content, spokesperson says
WhatsApp was not immediately available to comment.
Meta Platform’s messaging unit WhatsApp will likely be subject to tough online content rules targeting illegal and harmful content after it met the user threshold under this regulation, a European Commission spokesperson said on Friday.
WhatsApp had about 51.7 million average monthly active users of its WhatsApp Channels in the European Union in the first six months of 2025, above the 45-million-user threshold set out in the Digital Services Act (DSA), Reuters reported.
The DSA requires such large platforms to do more to tackle illegal and harmful content. Meta’s Facebook and Instagram, Google’s YouTube, TikTok, Temu and Microsoft’s Linkedin are some of the companies labelled as very large online platforms under the DSA subject to this requirement.
“So the objective for the Commission here is to check what is actually private messaging which doesn’t fall under the scope of the DSA and what are open channels that act more as a social media platform, this falls under the scope of the DSA,” Commission spokesperson Thomas Regnier told a daily news briefing.
“So here we would indeed designate potentially WhatsApp for WhatsApp channels and I can confirm that the Commission is actively looking into it and I wouldn’t exclude a future designation,” he said.
WhatsApp was not immediately available to comment.
Companies risk fines of as much as 6% of their global annual revenue for DSA violations.
-
Sport4 days agoAFC Futsal Asian Cup 2026: Final eight confirmed
-
Sport4 days agoAfghanistan in new kit for T20 World Cup warm-up against Scotland
-
Sport5 days agoIran see off spirited Afghanistan to finish top of Group D
-
Sport3 days agoJapan trumps Afghanistan 6-0 in AFC Futsal Asian Cup quarter-final
-
Sport2 days agoHosts and heavyweights advance as AFC Futsal Asian Cup reaches semifinals
-
International Sports4 days agoPakistan to boycott T20 World Cup group match against India
-
Sport4 days agoAfghanistan crush Scotland in ICC T20 World Cup warm-up
-
Latest News1 day agoTerrorist threat in Afghanistan must be taken seriously, China tells UNSC
