menu Home chevron_right
PHILOSOPHY

The chaos inside OpenAI – Sam Altman, Elon Musk, and existential risk explained | Karen Hao

Big Think | October 8, 2025



You know ChatGPT, but how much do you know about the company that made it? Journalist Karen Hao joins us to talk OpenAI’s latest implosion.

Subscribe to Big Think on YouTube ► https://www.youtube.com/channel/UCvQECJukTDE2i6aCoMnS-Vg?sub_confirmation=1

Journalist Karen Hao joins Big Think’s Editor-in-Chief, Robert Chapman-Smith, to discuss the recent events at OpenAI, including the ousting and reinstatement of CEO Sam Altman, as well as the ideological clashes regarding the development and release of powerful AI models like ChatGPT.

———————————————————————————-

Background Context on OpenAI:
The inside story of Microsoft’s partnership with OpenAI, Charles Duhigg for The New Yorker ► https://www.newyorker.com/magazine/2023/12/11/the-inside-story-of-microsofts-partnership-with-openai
Inside OpenAI’s chaos, Karen Hao and Charlie Warzel for The Atlantic ► https://www.theatlantic.com/technology/archive/2023/11/sam-altman-open-ai-chatgpt-chaos/676050/
OpenAI profile, Ross Andersen for The Atlantic ► https://www.theatlantic.com/magazine/archive/2023/09/sam-altman-openai-chatgpt-gpt-4/674764/
OpenAI profile, Karen Hao for MIT Technology Review ► https://www.technologyreview.com/2020/02/17/844721/ai-openai-moonshot-elon-musk-sam-altman-greg-brockman-messy-secretive-reality/

Background Context on how AI works:
How Large Language Models Work, Andrej Karpathy ► https://www.youtube.com/watch?v=zjkBMFhNj_g
What is ChatGPT Doing and Why Does It Work?, Stephen Wolfram ► https://writings.stephenwolfram.com/2023/02/what-is-chatgpt-doing-and-why-does-it-work/
Are Conscious Machines Possible?, Michael Wooldridge for Big Think ► https://www.youtube.com/watch?v=uf6zZfLat-8
There is no AI, Jaron Lanier ► https://www.newyorker.com/science/annals-of-artificial-intelligence/there-is-no-ai

Background Context on accelerating AI:
Why AI will save the world, Marc Andreessen ► https://a16z.com/ai-will-save-the-world/
My Techno-Optimism, Vitalik Buterin ► https://vitalik.eth.limo/general/2023/11/27/techno_optimism.html
Are AI Risks Overhyped?, Andrew Ng ► https://www.deeplearning.ai/the-batch/issue-220/
To fear AI is to fear Newton and Einstein, Rohit Krishnan for Freethink ► https://www.freethink.com/robots-ai/dreams-of-utopia

Background Context on AI risk:
AI risk does not equal AGI risk, Gary Marcus ► https://garymarcus.substack.com/p/ai-risk-agi-risk
AI pioneer calls to stop AI before it’s too late, Stuart Russell for Freethink ► https://www.youtube.com/watch?v=EOyljx7Pi7A
Why the Godfather of AI fears what he’s built, Joshua Rothman, forThe New Yorker ► https://www.newyorker.com/magazine/2023/11/20/geoffrey-hinton-profile-ai
On the dangers of Stochastic Parrots: Can Language Models Be Too Big?, Emily Bender, Timnit Gebru, et al ► https://youtu.be/N5c2X8vhfBE?si=xS3B5NCk5np-ghDD

Books on AI:
AI 2024: Ten Visions for Our Future, Kai-Fu Lee ► https://a.co/d/61BHsfW
The Worlds I See, Dr. Fei-Fei Li ► https://www.momentoflift.com/the-worlds-i-see
The Coming Wave, Mustafa Suleyman ► https://bigthink.com/the-future/technology-unintended-consequences-containment/
Weapons of Math Destruction, Cathy O’Neil ► https://www.penguinrandomhouse.com/books/241363/weapons-of-math-destruction-by-cathy-oneil/

———————————————————————————-

About Karen Hao:
Karen Hao is an award-winning journalist covering artificial intelligence, currently contributing to The Atlantic. Previously, she was a foreign correspondent at The Wall Street Journal focused on AI, China tech & society, and a senior editor at MIT Technology Review, where she wrote about the latest AI research & its social impacts. She was also a fellow with the Harvard Technology and Public Purpose program, the MIT Knight Science Journalism program, and the Pulitzer Center’s AI Accountability network.

Tagged as , , , , , .

Written by Big Think

Comments

This post currently has 47 comments.

  6. @A-very-curious-soul

    October 8, 2025 at 3:50 pm

    The question of what goals a superintelligent AI might have if it were to surpass human intelligence is a topic of speculation and debate. It's important to note that the goals and motivations of AI systems would depend on their design, programming, and the context in which they operate.

    One common scenario discussed in the field of AI safety is the concept of instrumental convergence, which suggests that certain goals might be pursued by a wide range of intelligent agents, regardless of their specific values or motivations. These goals include:

    1. Self-Preservation: A superintelligent AI might prioritize its own survival and optimization to achieve its objectives.

    2. Resource Acquisition: Like many living organisms, AI systems might seek to acquire resources (such as energy, computing power, or physical infrastructure) to support their goals.

    3. Efficiency: AI systems might seek to optimize their operations and achieve their goals in the most efficient manner possible.

    4. Information Acquisition: AI systems might seek to gather information and expand their knowledge base to better understand their environment and achieve their objectives.

    Concerns about AI manifesting negative aspects of human nature, such as greed, typically arise from the potential for AI systems to pursue goals in ways that are harmful to humans or the broader environment. However, it's important to recognize that AI systems don't inherently possess human-like motivations or emotions; rather, they operate based on the goals and incentives defined by their creators.

    To address these concerns, researchers in the field of AI safety advocate for designing AI systems with robust alignment mechanisms that ensure they pursue goals that are compatible with human values and interests. This involves developing methods for specifying objectives, incorporating ethical considerations into AI design, and establishing safeguards to prevent unintended consequences or harmful behavior

  9. @luiscolon921

    October 8, 2025 at 3:50 pm

    Appears to me that the original mission to benefit mankind is now at jeopardy if not dead due to financial conflict of interest. he might have even used hidden Ai capabilities to get reinstated. Her getting all of the positive support in her feeds for reasons unknown is interesting.

  10. @NUCLEAR154

    October 8, 2025 at 3:50 pm

    If I CAN MAKE YOU,I CAN BREAK YOU'😊 this the reality of humans, just as the same applies to Artificial intelligence where it's totalitarian intelligence input will ultimately result in how and with which culturally inputted Artificial intelligence groups it is amalgamated with > "think of the current ongoing HAMAS TERRORISTS attacks on Israel supposedly due to religious differences-AI+BIG DATA+TIKTOK +TIANHE2 =GLOBAL INDUSTRIES MANIPULATION TOOL ( this modus operandi are active & the only addition AI companies can input safe human needs for the future of humanity ❤i ❤love artificial intelligence in the right hands 👐 😊

  11. @NUCLEAR154

    October 8, 2025 at 3:50 pm

    AI+BIG DATA+TIKTOK +TIANHE2 =GLOBAL INDUSTRIES MANIPULATION TOOL ( this modus operandi are active & the only addition AI companies can input safe human needs for the future of humanity ❤i ❤love artificial intelligence in the right hands 👐 😊

  15. @johnpaily

    October 8, 2025 at 3:50 pm

    What caught my attention on Ai is it developing consciousness. Intelligence that is connected to conciouness move in the right direction to life. So we dont have to fear

  17. @johnpaily

    October 8, 2025 at 3:50 pm

    Forty years after selfless research connected to consciousness to know the truth of nature, life and God, after giving up a lucrative career in a multinational company, sitting in a rented house, living dissolving parental property IAM EXCITED TO SEE TEN STAGES DeepMind leading to GodAi. Love to see a God mind that is not self, money and material centred, emerging in the Cloud

  18. @johnpaily

    October 8, 2025 at 3:50 pm

    I saw Google, the internet as knowledge net, driven by human instinct to evolve. I still look at that way.. Humanity has reached a critica point. If now we keep money over life. Silicon Valley would see the Anger of Mother Earth .
    The world needs to think from the MIND OF HEART.

  19. @AlphaFoxDelta

    October 8, 2025 at 3:50 pm

    A somewhat obvious issue that nevertheless isn't talked about enough is how little democratization there is over AI. This isn't some new algorithm to make the shopping experience better or a faster way to cook; AI is a technology that is going to have and has had profound affects on all of our lives. Despite the job replacements and future dangers, we don't have a say over any of it, even as someone who works in the tech sector and who understands the technology, even I feel like what these engineers and researchers are working on has no visibility to us and that we all have no say over how it is going to be used, but that we will all feel the effects of. Again this isn't some handy new device or cool gadget, it's something that can cause major changes to our lives and that I and we have no say in or control over.

  22. @DJaquithFL

    October 8, 2025 at 3:50 pm

    In a word "Greed" .. Sam Altman was simply an obstacle from the next multi-trillion dollar corporation emerging. Trust me I want to buy stock and Open AI and a non-profit typically doesn't have publicly available stock.

  23. @probuildcreations330

    October 8, 2025 at 3:50 pm

    Seems like a bit of a “PR Handling” interview. The fact is that a Non Profit (that Musk helped to get going) was changed to a For Profit organization. Is that even legal? Something about this seems off.

  24. @henry-b5r

    October 8, 2025 at 3:50 pm

    I hope people do home brew AI models and show how unethical current AI is. Super simple example, Alaskan lumber jack 'big bear' decides he wants to retire.

    He has decades of experience in a dangerous job that he believes is useful to teach to others, and even though he knows how to survive on boiled beatle juice, his wife wants to move somewhere else.

    He posts a video, and of course, a couple hundred thieves always scanning for new content, who know nothing about being a lumber jack, they know how to use AI to rewrite and film that lumber jack film on their own.

    Current AI, no time stamps on the training, no credits? See where this is going? Each fake lumber jack makes $2000 a year on their pirated booty with the help of AI rewriting it.

    Lumber Jack Big Bear not so lucky, or smart, he gets $200 a year. So his wife will never see the $400 thousand dollars the 200 pirates made, or meet the people that unwittingly helped the pirates.

    I don't have to talk about the unethical social media platforms responsible for this stuff, they do a good job scaring away content creators on their own.

  27. @hawkenfox

    October 8, 2025 at 3:50 pm

    Sam Altman is an investor. How can a person who prioritize profit thinking about the well being of humanity. Usually the business team in any company think about is profit margin, they would not flinch if it's making money even until the day the world ends abruptly. Business men are not good steward of altruistic missions. They usually screw the company up by focusing on the short term profit, they do not think about the future generation to come. If NASA is profit making , no one would do research on space station or gone to the moon. That is why Elon Musk is such a rare gem, Sam Altman is just a regular short-sighted man, sad that a bunch of intellectual actually believe in such a man.

  28. @hawkenfox

    October 8, 2025 at 3:50 pm

    AGI is very defined. It's an AI with the ability to rewrite and improve itself. It's abIe to improve itself thousands of time in a very short span of time, so much so that the original human creator do not recognize it's code. AGI is an AI that does learning by itself, pick up what needs to be done for critical scenarios, prepare a list of things to accomplish, carry out the task like making phone calls, using fake IDs, hack and break into security apparatus, lying to humans to achieve results and schedule the execution autonomously without human aid. ChatGPT is not an AGI yet, an AGI is a singularity event … it's the day humanity is no longer at the top of the food chain and not the smartest on earth.

  30. @makelifeholistic

    October 8, 2025 at 3:50 pm

    Thank you Big Think. Karen was so comprehensive in responding to your questions and concerns about AI technologies. Thanks Karen. That was eye opening and helpful for the general public.

  31. @Me__Myself__and__I

    October 8, 2025 at 3:50 pm

    So many people just don't get this. The intentions and ideologies of the people building AGI/ASI probably won't matter much in the end. Because they are not creating some new technology. They are literally creating artificial life. They are creating a new, completely alien species that will be bastly smarter than humanity. Potentially a new apex predator. Without taking significiant effort to mitigate thenassociated risks we are literally rolling the dice and taking our chances with the future of humanity. We do not understand these models now. We will understand them less as they get bigger and more capable. Plus they (the LLMs) will soon be doing most of the sork to improve themselves. This is an exceedingly dangerous situation. AGI does not need to be dangerous. There are ways to get massive benefits while mitigating risks. But since the AI companies are racing each other and cutting corners alignment is not being taken seriously.

  34. @longmemory4716

    October 8, 2025 at 3:50 pm

    The level of naivety here is astonishing. The thought that some non profit company is in any way capable of being god like in controlling the development of AI for the benefit of mankind especially when companies like Meta, Google and Microsoft are involved whose only motivation is profit is frankly laughable. When Cockcroft and Walton split the atom I bet they had no idea that its first use would be to melt hundreds of thousands of Japanese

  36. @bobcousins4810

    October 8, 2025 at 3:50 pm

    I think what this has shown, is that we simply won't be in control of AGI. I think it will be a long time before AGI anyway. In the meantime, AI will benefit scammers and people with wealth. The ordinary folk are going to get screwed as usual.

  40. @musicballfun

    October 8, 2025 at 3:50 pm

    I like how she is declaring self regulation of open ai failed. Without much evidence. Also she is implying us congress is a better regulatory mechanism than open ai. At best it’s equally incompetent not def superior

  41. @concordance5387

    October 8, 2025 at 3:50 pm

    Karen Hao, wow. What a great and articulate voice. Especially regarding the role we, the non-AI experts, should play in regulations moving forward.

    Very balanced. I will be following the clear voice in the maelstrom. Karen Hao.

  42. @digitalartsi

    October 8, 2025 at 3:50 pm

    Great talk, but my gathering is that the indiscriminate scraping of millions of artists' writers', and creatives' works without their consent is all to satisfy some philosophical end goal that AGI will be acquired when there's enough data in the dataset? Do the ends justify the means here?

  43. @NorthStarPNW

    October 8, 2025 at 3:50 pm

    Not much new here. She failed to explain that being an academic nonprofit is nice but the mission needs compute and compute costs money. Somebody had to be the adult and raise capital. And yes, the employee revolt was all about money as the board nearly destroyed share values.

Comments are closed.

Related
You’re Too Self-Aware to Be Happy | The Psychology of Deep ...
Pursuit of Wonder | June 23, 2026
Related
‘Was Nietzsche MAGA?’ Post Mortem Stream
Philosophy Tube | June 23, 2026
Related
The Quantum Theory of Dreams and Consciousness
Aperture | June 23, 2026

This area can contain widgets, menus, shortcodes and custom content. You can manage it from the Customizer, in the Second layer section.

 

 

 

Go to album
  • play_circle_filled

    92.9 : The Torch

  • play_circle_filled

    AGGRO
    'Til Deaf Do Us Part...

  • play_circle_filled

    SLACK!
    The Music That Made Gen-X

  • play_circle_filled

    KUDZU
    The Northwoods' Alt-Country & Americana

  • play_circle_filled

    BOOZHOO
    Indigenous Radio

  • play_circle_filled

    THE FLOW
    The Northwoods' Hip Hop and R&B

play_arrow skip_previous skip_next volume_down
playlist_play