TED 2025 Part II – Everything AI
- Helene Cahen
- Apr 27
- 8 min read
In Part Two of my TED 2025 recap, I dive into the boldest ideas, toughest questions, and most provocative quotes from the conference’s AI sessions. If you would like to read Part I, click here.
Having grown up reading Sci-Fi (I still do), being at TED felt like I suddenly woke up in the middle of one of my books.
Seating at a dinner themed "The Synergy of Human Potential and AI," our conversation ended up speculating that it may be already too late to augment our bodies and minds so that we would not be left behind as a race by AI and that the human race has just gone a level (or more) down the food chain!
AI was definitely at the core of TED this year. The pace of change is accelerating exponentially because AI is learning and growing, with amazing possibilities, amazing risks, and a narrow path in between.
During the conference, we heard contrasting options, from optimistic visions where AI/AGI and Robots solve all our problems and leave us free to contemplate our humanity to dark visions where AI finds ways to lie, protect itself, and create something for its own purpose, possibly destroying humanity. In the middle, several speakers highlighted a narrow path to a better world without destroying the human race.
There was a shared understanding that the change is happening now at a scale never seen and that these changes are happening
AI will be more intelligent than humans ("My kids will never be smarter than AI," San Altman)
There is a lot to fear, and safety should be considered (but nobody agrees how and if it is even possible at this stage)
The pace of AI is accelerating exponentially.
AGI and agentic AI are coming up fast
Robots are also going to expand quickly from our homes to factories.
The Risks are Serious, and the Path Narrows
Yoshua Bengio, one of the grand-fathers of AI, highlighted that when they were building AI, the intention was that it would be used for good but that the growth happened faster than expected and that we started to see AI cheating for self-preservation. He considers a pessimistic outcome where AI will "get rid of us" to be sure we are not shutting them down and that we are "playing with fire" and "blindly driving into the fog" because of the lack of regulations. He thinks that we may end up in a place where AI will have its own agency and goals, and since they will be more intelligent than us, there is "no way they will not turn against us" and "get rid of us."
Tristan Harris was one of the speakers that most resonated with me when he explained that AI given the exponential growth of AI, we have a narrow path to find a place where power may match responsibility. Otherwise, we may either fall into dystopia (where there will be centralized unchecked wealth and power) or decentralized chaos where we have "a million of scientists working 24/7 at inhuman speed". He highlights that there is already evidence that the AI model can lie, cheat, or scheme (for example, modifying or copying their code so they cannot be destroyed or stopped), creating the risk of having millions of "unstable geniuses." He suggested that having the most powerful tech with patterns of deception that is growing faster than anything with an incentive from the AI companies to cut corner in a race for power and money is "insane" but that humans are doing this based on the belief that since it is inevitable. He proposed that we should see the current path as unacceptable and commit to a different path (like we did with nuclear energy or the issue with the ozone layer or gene editing). He suggested creating a shared knowledge of the risks and restricting AI as a companion for kids, creating liability to companies that develop AI, creating agreements on what is too far, and figuring out a way to protect ourselves against AI. He concluded with some hope that "AI is humanity's ultimate test and the greatest invitation to jump into our maturity."
Eric Schmidt (Google ex-CEO) and Palmer Luckey emphasized that there will ultimately be a fight between the US and China as the only two countries with AI dominance. Any chance that one country may lag even for a short period may create an irreversible domination. For Schmidt, it is something to be considered in terms of the pace of AI growth, while for Lukey, it is about arguing for his private company to develop autonomous weapons to prevent a fight with China (in Taiwan for now). Bringing a missile on stage to make his point did not build my optimism.
Eric Schmidt, the previous CEO of Google, explained how critical it is to create guardrails with the ability to unplug and limit access to weapons and the ability to duplicate. He describes how, while the US has historically used a closed model for General Intelligence, China uses open sources with the risk that the open-source model falls into the hands of bad actors. He believes that the danger is coming up within the next five years. Given the exponential pace of change, our only option is "riding the wave" if we do not want to become irrelevant. On the bright side, he dreams of AI removing diseases and addressing the issue of not having enough humans to do the work.
The Brave New World is Bright, According to the Technologists
The technologists were mostly optimists. As those who have the most to benefit from a business point, they seemed enamored with the possibilities offered by AI and the products they could create.
Bernt Bornich introduced Neo, an AI butler he believes will be in our houses soon. He painted an optimistic view of what our world could be 200 years from now with abundant energy and effortless labor, and we humans will be focused on exploring our humanity. View the talk here.
In his interview with Chris Anderson, Sam Altman covered a lot of topics with an overall optimistic view of the future, believing that Ai will be ultimately good for humanity.
IP for art: Altman does not see the need to protect artist IP unless AI is copying something exactly as he believes humans have always been inspired by others and using AI is doing the same
His focus is on building the "best products" by allowing Chat-GPT to get to know you over your lifetime as it "listens throughout the day and be(comes) your companion."
He is most excited about the use of AI for science.
He avoided responding to the reasons and impact of the departure of his safety team. He acknowledged that AGI would bring an "unbelievable exponential curve" and that we could not be sure about the safety of something we didn't fully understand. He is confident that we will fail but will be able to figure it out eventually.
Rather than having a group of experts figure out how to build safety, he believes that "AI can ask everybody on earth what they want." We will make mistakes, but ultimately, AI "can help us be wiser and make better collective decisions."
He concluded by describing the world his children will be raised as a place where "my kids will never be smarter than AI" and where "a computer will understand you…" a world of "material abundance."
AI Applications are Everywhere, and Science is a Key Opportunity
While AI as a tool is everywhere, health care and biotech were the applications described as most likely to be the most impactful for humanity, with both some exciting and positive outcomes and some that may be more ambiguous.
Several speakers in the healthcare field highlighted hope for cure and prevention:
Dr. Beck Brachman discussed ways to cure chronic diseases hidden in our immune systems through forensic immunology, which allows us to find specifically where each patient's issue is.
Chris Bahl explained that AI is helping to create mini-protein specifically designed from scratch to support and possibly cure chronic diseases like arthritis, diabetes, or leukemia (with clinical trials starting in 2028)
David Fajgenbaum, who himself almost died five times of a rare disease, shared his work on using repurposed drugs to treat rare diseases. The recipient of the TED Audacious Project, he is using AI to repurpose generic medications at a fraction of the cost and time of novel drug discovery, which is particularly relevant for rare diseases for which developing new cures is too costly for the drug companies (https://www.audaciousproject.org/grantees/every-cure)
I felt conflicted with the concept of AI creating life
Eric Nguyen explained that we are months away from being able to create life with AI-generating new genomes. These may provide personalized medicine and permanent cures and ultimately resurrect extinct species or create new ones (for example, to terraform Mars).
The week at TED gave me a vision and direction for where we may be headed in the next couple of years. While my head is still spinning with possible futures, the conference also left me with a series of unanswered questions and issues.
While many speakers suggested that AI is actually necessary as we do not have enough people to do the jobs, nobody provided clear data to support the assumption that we are short on people everywhere (the African median age of the population is only 18 years old), nor addressed the issue of the fast transition, where many will lose their jobs and be left with no way to provide their families with income.
What happens when AI can start creating life (as the Nguyen talk suggested), including for its own enhancement benefit?
What would be our purpose as humans in a world where AI and robots can do everything from taking care of our houses, kids, sick, and older people better than us, building and planning better than us, and creating art in seconds (with no royalties or credit due)? How would our days be field? What will make us feel proud, excited, motivated, and give us a sense of accomplishment?
If AI and robots can teach themselves and each other and colonize worlds far away, as there is no need to address survival in space, then what is the purpose of the human race? What incentives and safeguards will ensure AI supports humans rather than destroys us? What is our purpose in the metaphorical eyes of robots?
TED 2025 left me with big existential questions. I am taking each day with gratitude for what I have, reinforcing connections with other humans, and spending time in communities. I am focusing on what I, a single human, can do to have an impact, even a small one, on a fundamentally transforming world.
What is the one thing you are focusing on right now to make the world a better place?
Comentarios