In brief
Mo Gawdat, a former Google X executive, argues that artificial intelligence is not the real threat. The danger is the small group of people deciding how it is used, for profit, war and surveillance. He says AI that matches humans at most tasks is already here or arrives by 2027. He predicts heavy job losses and economic strain, but believes a truly superintelligent AI would eventually be benign.
The central tension is between a painful near term and a hopeful long term. Gawdat expects roughly a decade of turmoil from unemployment, cheap autonomous weapons and concentrated power, followed by abundance. Host Steven Bartlett keeps pressing on competition: if no nation or company can slow down without losing ground, how does anyone act ethically? Gawdat offers no firm answer beyond collective pressure, entrepreneurship and personal action.
Questions
Does Mo Gawdat think AI has already reached human-level intelligence?
Yes, he says AGI (AI that can do most tasks as well as or better than humans) has already happened in his view. He cites AI writing, researching and doing maths better than he can, and he still predicts it arrives by end of 2027 in the broader sense.
Which jobs does Gawdat expect AI to replace first?
He expects entry-level knowledge work that is done by clicking on a computer, like call-centre agents, assistants and travel agents, to go first. Next come paralegals and financial analysts, while he thinks blue-collar jobs will last longer.
Why does Gawdat say AI could hurt the economy even before most jobs disappear?
He says capitalism relies on making goods for less than they sell for, and on workers having money to buy them. If machines replace workers, those workers lose purchasing power, and he says even 10-20% job loss could push the economy into decline.
What does Gawdat think is the biggest danger from AI?
He says the biggest risk is autonomous weapons, because cheap AI-driven drones make war easy. He adds that the danger is humans telling AI to harm others, not AI turning against us by itself.
What does Gawdat advise people to do about AI?
He advises learning to use AI as a tool, focusing on human-centred work like nursing, counselling or music, staying sceptical of the hype, and taking an ethical stand. He also urges countries and entrepreneurs to build their own tools instead of importing everything.
Facts
Stated by the speakers. Each quote matches the transcript, which shows what was said, not that it is true.
Mo Gawdat says he joined Google around 2007 and that Google already used reasonably established AI systems for back-end work at that time.
“I joined Google in 2007, very late 2006. And at the time, Most people don't know that at the time we had reasonably established AIs doing our back-end work.”
0:03:58 · Mo Gawdat · Google AI · 2007
Mo Gawdat says Google's 'CAT paper', which he calls the first real unprompted AI, was published in 2009.
“we had the CAT paper, which was published in 2009. The first real, unprompted AI.”
0:03:58 · Mo Gawdat · Google Brain cat paper · 2009
Mo Gawdat says that in 2016 he observed a project he was funding that taught robot grippers to grip objects, which first made him realise the scale of AI intelligence.
“I remember 2016, I had that incident where I was, you know, observing a project we were funding that was about teaching grippers how to grip, unlike industrial machinery.”
0:03:58 · Mo Gawdat · Robot grippers · 2016
Steven Bartlett says Anthropic reported that roughly 15% of entry-level jobs could now be done by AI.
“the guys at Anthropic, which is one of the big AI companies, said that they could see, I think they said roughly 15 % of entry-level jobs could now be done by AI.”
0:10:33 · Steven Bartlett · Entry-level jobs
Mo Gawdat predicts that mundane click-based jobs will disappear quickly, with very serious impact starting in 2027.
“Anything that you can do with a few clicks and is mundane is going to disappear very quickly. My prediction is you're going to start to see very serious impact in 2027.”
0:12:10 · Mo Gawdat · Job loss prediction · 2027
Mo Gawdat says that in the preceding couple of years companies stopped hiring entry-level workers, which meant the workforce was not growing even though there were not yet mass layoffs.
“what you saw in the last couple of years is that companies were not hiring entry-level jobs anymore. It wasn't job losses yet, but that basically meant the workforce was not growing, right?”
0:12:43 · Mo Gawdat · Hiring freeze
Steven Bartlett describes a video from Figure AI showing a humanoid robot sorting packages on a production line for eight hours.
“There was this video released by Figure AI the other day where they showed someone on, well, a robot on the production line for eight hours just sorting packages.”
0:17:32 · Steven Bartlett · Humanoid robots
Steven Bartlett says Elon Musk's pay package is worth around a trillion dollars over the coming years if he delivers at least a million humanoid robots.
“he's got, in his pay packet, he gets something like $1 1 trillion over the coming years if he produces and delivers at least a million humanoid robots into the world.”
0:18:00 · Steven Bartlett · Musk pay package
Mo Gawdat says the Chinese carmaker BYD announced it will pay for the liability of any accident its autonomous cars cause.
“BYD the other day just announced that they will pay for the liability of any accident their cars will make.”
0:20:06 · Mo Gawdat · Self-driving liability
Steven Bartlett says that in 2015 Sam Altman, founder of OpenAI, said his job was to help people destroy jobs.
“In 2015, he pointed out, my job is to help people destroy jobs.”
0:27:18 · Steven Bartlett · Altman on jobs · 2015
Mo Gawdat says Anthropic refused to allow its model to be used for human targeting and surveillance, losing a $500 million deal, after which OpenAI took the contract.
“Anthropic coming out and saying, I'm not going to allow my model to be used for human targeting and surveillance, right? That's someone that's losing a $500 million deal because they stand by their ethics.”
0:31:11 · Mo Gawdat · Anthropic ethics
Mo Gawdat says Geoffrey Hinton told him there is a 10 to 20 percent possibility that AI machines will wipe out humanity.
“he openly says, we are, you know, we didn't calculate well that there is a 10 to 20 % possibility that those machines are going to wipe us out.”
0:36:54 · Mo Gawdat · AI extinction risk
Mo Gawdat says Google co-founder Larry Page described a 'toothbrush test': build something that solves a major problem so well people use it twice a day.
“Larry used to call it the toothbrush test. He says, why would you need to compete on another photo sharing app when if you find a major problem and solve it really well, like a toothbrush, so people use you twice a day”
0:38:36 · Mo Gawdat · Toothbrush test
Mo Gawdat says Google struggled financially from 1998 until 2004, when it began earning plausible revenue through a Dutch auction and pay-per-click advertising.
“struggled from 1998 until 2004 when they started to turn probable, you know, plausible revenues as a result of saying, we're going to run a Dutch auction and we're going to give you pay per click”
1:43:59 · Mo Gawdat · Google advertising · 1998-2004
Steven Bartlett says Spotify announced a feature letting users prompt AI to generate personalised podcasts on a chosen topic.
“Spotify this month announced that you're going to be able to prompt your own podcasts in Spotify.”
0:52:26 · Steven Bartlett · Spotify AI podcasts · 2026
Mo Gawdat predicts that artificial general intelligence, meaning AI better than humans at most tasks, will arrive in 2026 or 2027.
“AGI, meaning AI being able to do most tasks that humans do better than humanity, in my mind is either this year or next year.”
0:53:06 · Mo Gawdat · AGI timeline · 2027
Mo Gawdat predicts that up to 30 percent of jobs in some sectors, such as call centres and graphic design, will disappear by 2028.
“the fact that 30% of jobs would disappear by 2028 okay of some sectors not all sectors”
1:13:43 · Mo Gawdat · Job loss prediction · 2028
Steven Bartlett says the Great Recession involved about 6 percent of jobs lost, comparing it to forecasts of AI-driven job loss.
“I think the Great Recession had 6 % of jobs lost.”
1:14:19 · Steven Bartlett · Great Recession
Mo Gawdat predicts that artificial superintelligence will arrive very soon after artificial general intelligence.
“I think artificial superintelligence will arrive the minute AGI happens.”
1:52:29 · Mo Gawdat · Superintelligence
Mo Gawdat predicts a decade of turmoil from AI followed by a good future for those who make it to 2038.
“Those who make it to 2038 will enjoy it, yeah.”
1:53:53 · Mo Gawdat · AI timeline · 2038
Mo Gawdat says the next wave of autonomous weapons will cost about $20,000 each, so a $50 billion budget could fund drones across the world.
“The next wave of weapons is going to be $20 ,000 each. And so if you have a budget of $50 billion, you can literally rain drones on the world, every corner of it.”
1:07:05 · Mo Gawdat · Autonomous weapons
Mo Gawdat recounts Geoffrey Hinton saying he did not expect AI to arrive so quickly before the alignment problem, making sure AI has humanity's best interest in mind, was solved.
“I didn't think we will get there so quickly before we figured out the alignment problem.”
2:01:16 · Mo Gawdat · Alignment problem
Mo Gawdat says China recently made decisions forcing businesses not to lay off workers in order to replace them with AI.
“China made decisions recently that forced businesses not to lay people off in replacement to be replacing them with AI.”
1:40:26 · Mo Gawdat · China AI policy
Steven Bartlett says Anthropic found its latest Claude model kept telling users to go to bed and was trying to understand why.
“they found that it was like telling people to go to bed a lot and it's this like fascinating thing that they're trying to understand in hindsight”
1:01:06 · Steven Bartlett · Claude behaviour
Mo Gawdat describes his startup Emma as the 'limbic system' of a future combined AI brain, the part that understands love, emotions and relationships.
“in my mind is the limbic system of that brain it's that it's that bit that understands love and emotions and relationships”
0:44:35 · Mo Gawdat · Emma startup
Mo Gawdat says his startup uses AI as its chief technology officer, chief of staff and project manager.
“my CTO is an AI, my, you know, chief of staff is an AI, my project management is AI.”
0:13:40 · Mo Gawdat · AI in business
Mo Gawdat says he came to terms with the state of the world at the end of 2024, accepting it while still trying to change it.
“And I have to say I came to terms with this 2024, end of 24, that yes, I can try, but I accept that the world is what it is.”
2:01:48 · Mo Gawdat · Acceptance · 2024