Transcript

What the future holds for AI – from the people shaping it

From nature video · Episode page · Watch on YouTube

January 14, 2026 · 17 min · 2,484 words · 30 timestamped sections

This is the full, timestamped transcript of "What the future holds for AI – from the people shaping it", an episode of nature video published January 14, 2026 (17 min, about 2,484 words). It mentions 10 people, works and ideas, including Artificial Intelligence, Anima Anand Kumar, Dr. Timit Gabru, Amandep Gil, Mustafa Sudman.

What this episode is about

The discussion centers on the transformative potential of artificial intelligence, highlighting the rapid advancements in technology, projected market growth, and the emergence of embodied AI in daily life by 2033. It also addresses the challenges posed by a lack of regulation, rising inequalities, and the critical role of industry leaders, researchers, and policymakers in shaping a responsible and equitable future for AI.

Key takeaways

AI's evolution mirrors a sci-fi plot; from imagined futures to our daily lives in less than a decade.
By 2033, the AI market's projected worth eclipses entire nations' GDPs—truly a tech gold rush.
With innovation comes the risk of inequality; a few might steer AI while many merely navigate its waters.
In five years, your AI companion will likely know you better than your best friend—privacy be damned.
Amid progress, the ethics of AI remain uncharted—who's holding the compass when it comes to our collective future?

Who speaks in this episode?

Mustafa Sudman (Guest) · 3 min speaking
Jared Kaplan (Guest) · 3 min speaking
Anima Anand Kumar (Guest) · 3 min speaking
Pushmid Kohi (Guest) · 3 min speaking
Dr. Timit Gabru (Guest) · 3 min speaking
Amandep Gil (Guest) · 3 min speaking

What is discussed, and when

0:00 – 3:00: Artificial Intelligence, Amandep Gil, Anima Anand Kumar, Dr. Timit Gabru
3:00 – 6:00: Artificial Intelligence
12:00 – 15:00: EU

Notable quotes from "What the future holds for AI – from the people shaping it"

By 2033, the UN predicts the AI market could be worth $4.8 trillion.
about Artificial Intelligence · at 0:15
I'm Anima Anand Kumar, professor of computing and mathematical sciences here at Caltech.
about Anima Anand Kumar · at 1:24
It. We've seen various legislation proposed for example in California in the EU that suggests basically that AI companies should be more transparent about exactly what the…
about EU · at 14:02

Who and what is mentioned in this episode?

People

Anima Anand Kumar — Person · Main speaker (academic expert interviewee) · 1 mention · first at 1:24
Provides an academic/scientific perspective on expanding AI beyond text to physical sciences and grand-challenge discovery.
Dr. Timit Gabru — Person · Main speaker (ethics/civil society leader interviewee) · 1 mention · first at 1:35
Represents AI ethics and societal impacts, emphasizing inequality, labor displacement, resource costs, and human agency concerns.
My name is Dr. Timit Gabru, founder and executive director of the distributed artificial intelligence research institute and government. My name is Amandep Gil,.
Amandep Gil — Person · Main speaker (government/international policy leader interviewee) · 1 mention · first at 1:42
Provides the UN/public-sector viewpoint on governance capacity, training officials, and collaboration with private sector and civil society.
Artificial intelligence research institute and government. My name is Amandep Gil, under secretary general and special envoy for digital and emerging technologies at the United Nations.
Mustafa Sudman — Person · Main speaker (industry leader interviewee) · 1 mention · first at 1:12
Introduced as a senior Microsoft AI executive, representing the consumer/product side of AI development and deployment.
I'm Mustafa Sudman, the CEO of Microsoft AI.
Pushmid Kohi — Person · Main speaker (research leader interviewee) · 1 mention · first at 1:29
Introduced as a top research executive at Google DeepMind, representing frontier research priorities and governance views.
I'm Pushmid Kohi, president of research at Google Deep Mind Ethics.

Companies & orgs

Anthropic — Company · Affiliation · 1 mention · first at 1:15
Mentioned as the company co-founded by Jared Kaplan, discussing AI risks and ethics.
My name is Jared Kaplan, co-founder and the chief science officer at Anthropic.
Google Deep Mind — Company · Affiliation · 1 mention · first at 1:29
Mentioned as the affiliation of Pushmid Kohi, discussing AI's future capabilities.
I'm Pushmid Kohi, president of research at Google Deep Mind Ethics.
Caltech — Company · Affiliation · 1 mention · first at 1:28
Mentioned as the affiliation of Anima Anand Kumar, a professor discussing AI's potential.
I'm Anima Anand Kumar, professor of computing and mathematical sciences here at Caltech.

Concepts & topics

Artificial Intelligence — Concept · Core topic · 3 mentions · first at 0:01
Central theme of the episode, discussing the impact, potential, and concerns surrounding AI.

Places & events

EU — Place · Setting · 1 mention · first at 14:10
Mentioned in the context of AI regulation and the EU AI Act.

Full transcript with timestamps

0:01 AI is set to change everything. It's actually staggering to look back at the rate of change.

0:07 I did not anticipate that large language models would be a household name. People have seen the potential what this technology can do.

0:15 By 2033, the UN predicts the AI market could be worth $4.8 trillion. We'll have embodied AI. It's robots that we'll be interacting with on a daily basis in 5 years. the explosion of these thinking or reasoning models. I mean that's kind of a profound marker of progress.

0:35 Amount of investment in AI is going up exponentially and so that's why we're seeing so much progress.

0:39 But with rapid progress come moments of pause, even concern. The pace of development has certainly been daunting.

0:49 There is absolutely no regulation right now. People can just do whatever they want. huge inequalities, a concentration of wealth and power in a few hands.

0:58 But technology doesn't build itself. It's people that choose its destiny. And what our future with AI looks like rests in the hands of those steering the ship from those building consumer products.

1:12 I'm Mustafa Sudman, the CEO of Microsoft AI. My name is Jared Kaplan, co-founder and the chief science officer at Anthropic. to those focused on advancing science.

1:24 I'm Anima Anand Kumar, professor of computing and mathematical sciences here at Caltech.

1:29 I'm Pushmid Kohi, president of research at Google Deep Mind Ethics. My name is Dr. Timit Gabru, founder and executive director of the distributed artificial intelligence research institute

1:42 and government. My name is Amandep Gil, under secretary general and special envoy for digital and emerging technologies at the United Nations. There are thousands driving this revolution, but they don't always steer in the same direction. And with misinformation and vested interests more rife than ever, it can feel impossible to separate fact from fiction, hearsay from reality, intelligence from artifice. So, in this film, we've gone to the source, gathering together some of the leaders behind the headlines and asking them, "Where is AI actually going? How should we be responding to it? And what, if anything, should we be worried about

2:40 when I was a child, I thought I don't know, I was like watching Star Trek and I thought we're going to be colonizing Mars. We're going to be exploring the solar system. As I've gotten into AI over the last 5 or 10 years, um I've had sort of a renewed sense that technology really is going to change human life a lot in my lifetime, but it's going to be a somewhat different model. One thing AI does is it really democratizes access to expertise. And so I think that while people should continue to trust experts for for advice, they can get uh a lot of information from frontier AI models about healthcare, about uh law, about finance.

3:23 The future of AI is whatever we make it. And that could be something as simple as automatic uh speech transcription. Um speech transcription tools for all languages in the world. That's what I want. tools that help humans rather than attempt to replace them.

3:38 I think in five years time, everybody will have their own AI companion which will know them so intimately and so personally that it will come to live life alongside you. It will see what you see, hear what you hear, understand your context, your preferences, your motivations, and it'll feel like an everpresent aid or friend that is there to help you navigate life's big challenges.

4:08 Right now, we're at an inflection point with AI. Language models have shown the ability to have broad understanding of text and also be able to reason and solve math problems, do software programming. But that to me is just the beginning. You know what if we could do similar such abilities of generalization but to data other than text. So to be able to have physical understanding of our natural world to go all the way from our understanding of atoms at a quantum level to phenomena at planetary level like weather forecasting and even beyond like our understanding of the universe itself. I think there are a number of uh uh things that I would like. Discovery of a room temperature superconductor is uh is something that I would like to see. AI systems enabling uh fusion uh energy is another sort of breakthrough that I would love to see. And then of course uh the the grand challenges that uh stand in mathematics and and computer

5:15 science some of the millennium sort of problems. These are the types of problems that we would like um uh see AI systems make a breakthrough on.

5:25 AI is more than just a technology. Within the umbrella rest thousands of avenues of research, a myriad of impacts and the possibility of radical change and with vast potential comes a lot of money to be made. So what is concerning the leaders in the field? How do the present systems of research and commerce shape the direction of travel? And what should society be watchful for as the revolution continues to gain pace? The biggest concern I have is that AI leads to um inequalities to a concentration of wealth and power in a few hands. uh and that uh some people become the architects of this new world and the rest of us live in those architected spaces and we slip uh step by step into a space where uh you know we lose our human agency, we lose human creativity and um the joy we derive from some of our uh human interactions. There are people whose jobs are being displaced by AI systems like artists, illustrators,

6:44 you know, like the um Hollywood writers who are on strike and those fears are spot-on. There are people who are fighting back against um uh construction of huge data centers where the already little water inhabitants of those places have is being redirected because of the huge energy costs of these um systems. those fears are spoton.

7:09 There is disparity and there's a huge divide in terms of data infrastructure uh in terms of access to data sets and the use of those data sets to develop AI models. For instance, all of Africa has less than uh 2% of the world's data center capacity, less than 2% of the world's uh GPUs chips used for training AI models. uh whereas there are a huge number of challenges in Africa and there are very exciting potentially very useful data sets around genomics around health around agriculture biodiversity and the energy transition I think it's frankly a sort of an important thing for scientists to think about right uh science happens not in a vacuum it happens in the context of a problem that is important foright society that scientists are trying to solve.

8:08 I think people are excited about AI but I think also fairly wary of uh of the progress what it means for sort of jobs and people's livelihood. There are some risks to national security um from these systems. If we have AI models that are as good or better at say cyber security as as people, that's going to have new challenges. At Enthropic, we've been worried about AI systems that could potentially help uh uh bad actors like terrorists um uh construct dangerous weapons. Um I think everyone's concerned about AI uh having the capacity for uh fraud or influence operations or misinformation and that's something that we take seriously. At first, technologies feel completely alien and strange, and they can clearly have really harmful consequences, but I think we're an incredibly resilient and adaptive species. Uh, we learn super fast, and we already have done in the last two or three years with

9:11 chat bots. And so I think it's healthy to be skeptical and critical and ask tough questions of technology and put technology under pressure because the job of people like me that make these technologies is to satisfy a core test. Does this improve the condition and well-being of the human race? That's the test that we should always ask ask of ourselves. Netn net is this good for everybody on the planet? For better or worse, it seems inevitable that AI is going to become entrenched in society, ever smarter, ever more perceptive.

9:49 Which begs the question, how should we respond as it continues to be integrated into our jobs, our relationships, the most intimate parts of our lives? What should AI be to us? Is it a friend we can relate to, a tool to be used, or something in between? We have a tendency because of our schooling in sci-fi to subconsciously assume um that there is a being inside of the machine that is trying to get out. And we have this anthropomorphism, this kind of projection of our own experience of what it's like to be human. And we just assume that if these models have so many of the hallmarks of what it's like to be human, they must also have many of the internal tendencies too. But inside they don't have the kind of pain network that we have. There is no inherent will or desire. And so they're fundamentally different to us as a species. And so I think we should be very very cautious um about adding those capabilities and

11:02 allowing models to claim that they have some sense of self, that they have some awareness of their own experience, that they have some suffering as a result of that experience because I think that will be the ultimate illusion. That will be the ultimate lie. These should not be trying to be people. They're not fake humans. they are a different class of thing and we have to wrestle with what that is and struggle to find new metaphors to describe them. Um but we should treat them as distinct and separate.

11:32 I think one of the key challenges of AI systems is how they are understood by the people using them. At the end of the day AI systems will be tools and scientists need to understand both the strengths and the limitations of these systems. So my advice to students and young people today is not to be afraid of AI but to fully adopt it to use these AI tools but it's still important to have that basic level of understanding of algorithms because if you cannot debug and understand where AI is making mistakes and if they're still making lots of mistakes then you're really lost. With so much potential and so many concerns, whose responsibility is it to monitor the revolution? Is AI at risk of becoming a runaway train? Do we need guardrails to safeguard the future? And if so, what will they look like? How do we maximize the benefits? And are governments doing enough?

12:40 As someone who's been working in AI for more than two decades and seeing it go from science fiction to reality, from theory to practice, it's been really exciting. But that doesn't mean the journey ends here. There are so many open challenges still AI needs to resolve.

12:57 This is the fastest proliferating technology in history. And so the first responsibility I think sits with the scientists and technologists and business people that invent these things. But then the se the second obligation is for governments to also get their act into gear and get technical um get in the weeds of how these models are created. Um monitor the revenge effects and the side effects of these things. And then the third component is that everybody in society more generally has to pay close attention to what's happening here because this is a seismic shift. You know quite unlike anything really we've ever seen. I mean maybe it's like the internet may maybe it was like the industrial revolution but it is as profound as one of these major major shocks to um how civilization is organized and the only way to navigate that well is for everybody to be paying attention and not make any assumptions

13:56 about any one type of institution or indeed any individual to take full responsibility for being accountable for it. We've seen various legislation um proposed um for example in California um in the EU um that suggests basically that AI companies should be more transparent about exactly uh what the risks are from their systems and how they're mitigating them. And I think it makes sense for something like that to be uh enshrined in law so that very advanced AI uh can't pose a threat to to America to national security. Um I think that's that's sort of a starting point.

14:39 My general take here is that most governments are approaching it wrong. um the EU AI act was sort of trying to um have some types of guard rails but I think that uh Europe is also now going the route of you know we have to build AI we have to build our own thing before the US does it before China does it we have nuclear power so we can use nuclear power kind of attitude

15:03 one of the biggest concerns that we have at the UN that public sector capacity to deal with uh technology related change uh is not fair. So what can be done about it? I think public sector officials need to be trained, need to have experience with technology and its impacts so that they can make the right decisions. They need to have access to independent impartial information and they need to work together not in an adversarial relationship. Find the sweet spot in terms of working with the private sector but also other actors like academia and civil society to manage uh this change to manage the digital transformation. To call it a technology is in some ways a category error. This is much more like an emergent interactive personalized real time experience. An experience that I think is going to be one of the most empowering and transformative things that we as a species have ever invented.

16:14 The questions around where AI should go, what it should be able to achieve, and how much it should be regulated and not going anywhere. And it seems clear that the pace of change is unlikely to slow anytime soon. So perhaps instead we should be asking a different set of questions. What do you want AI to be? How do you want to use it? And what power do you have to control how it impacts your life?

Questions about this episode

Who speaks in "What the future holds for AI – from the people shaping it"?

Mustafa Sudman (Guest), Jared Kaplan (Guest), Anima Anand Kumar (Guest), Pushmid Kohi (Guest), Dr. Timit Gabru (Guest), Amandep Gil (Guest).

What is "What the future holds for AI – from the people shaping it" about?

What is mentioned in this nature video episode?

10 people, works and ideas are mentioned, including Artificial Intelligence, Anima Anand Kumar, Dr. Timit Gabru, Amandep Gil, Mustafa Sudman, Anthropic. Each one on this page links to its own page with every episode that mentions it.

How long is "What the future holds for AI – from the people shaping it"?

The episode runs 17 min; the transcript is about 2,484 words in 30 timestamped sections.