AI

Sam Harris discusses recent developments in AI and his concerns about AI risk with Paul Bloom

Sam Harris expresses concern about the rapid adoption of GPT-4 and its implications for AI risk
The suddenness with which AI has improved and surpassed landmarks set by AI safety experts is alarming
The development of powerful AI without full understanding of its implications has magnified the near-term risk of unintended consequences and potential malicious uses of narrow AI
The AI arms race and gold rush may lead to undesirable outcomes, despite everyone wanting to avoid such outcomes
Harris worries about the inundation of fake information and lies on the internet due to the increased use of AI, which could render it unusable
The best case scenario for AI still looks uncanny, leaving many questions unanswered about our relationship to information
Paul Bloom has written a new book, "Psych: The Story of the Human Mind," which covers the full sweep of what we know or think we know about the human mind
Bloom believes that we don't have a theory of the human mind and that the mind is many things, making it difficult to summarize what we know
We know that the mind is the brain, and that a lot of our mental life is the product of natural selection and cultural evolution
We know that a lot of our mental life is inaccessible to our consciousness, and that we are often conflicted beings who can fool ourselves
Harris and Bloom discuss recent developments in AI and the tension between misinformation and free speech
Bloom believes that we can learn a lot more about human life through good novels, films, and TV series than through psychological researchSam Harris on AI, Consciousness, and Leaving Twitter
Sam Harris expresses surprise and concern about the development of AI, specifically GPT-4 and its capabilities.
He admits to dismissing statistical attempts to model the human mind in his book, but recent events have made him reconsider.
Harris acknowledges the success of models like GPT-4 but questions whether it tells us much about how our minds work.
He distinguishes between intelligence and consciousness as separate questions and their ethical implications.
Harris worries about the near-term chaos and harm of AI that falls short of AGI, particularly in terms of producing misinformation.
He predicts a world where most of what is online is fake, leading to greater siloing of information and a greater role for gatekeepers.
Harris explains why he left Twitter, citing how it was making his life worse and distorting his perception of other people.
He notes the difference in his life after leaving Twitter and how it feels like a pre-internet existence.
Harris acknowledges spending a lot of time online but feels something has been stripped out of his life.
He suggests spending less time dealing with strangers and notes how social media can bring out the worst in people.
Harris admits to being tempted to follow his lead but acknowledges the benefits of Twitter in terms of staying informed.
He concludes by stating how he was hooked on Twitter for years, following hundreds of smart and creative people who surfaced interesting articles and ideas.Sam Harris on the Distorted Sense of Reality on Social Media
Sam Harris has defected from using Twitter due to the distorted sense of reality it creates.
Twitter and other social media sites have a business model that relies on gaining people's attention algorithmically to maximize ad revenue.
The privileging of misinformation and outrage by algorithms is a problem.
Harris is worried about our collective ability to have a fact-based discussion about anything due to the AI component and the misinformation component of social media.
AI algorithms are becoming geared for individuals, which can lead to losing time and becoming disconnected from the outside world.
The YouTube algorithm, which suggests content based on preferences, can lead to a loss of time and addiction.
Twitter and other social media sites can lead to a distorted sense of other people, which can change our opinions of them negatively.
It is becoming harder to read books due to the fast-paced nature of social media.
The outrage age machine has no off button and the pace of our engagement with the story of the day makes it difficult to solve long-standing problems.
The more we solve the problems of social media, the more boring it might get.
There is a concern about the polarizing nature of social media and the difficulty in converging on a shared set of truth claims.
There is no natural market solution to the problems of social media, and all of the proposed solutions have potential drawbacks.Discussion on the Human Mind and Truth in the Age of Misinformation
The euphoric feeling of connecting with people and maintaining reputation can be addictive and unhealthy, but when done properly, it can be a real plus of these algorithms.
Misrepresentation of views is the worst thing about reputational maintenance and caring about what others think.
The problem of misinformation at scale is worrying as it can lead to an epistemological crisis where people stop caring about the truth.
Trump started a trend where people don't care about the truth, which is a major problem in the world we live in.
The difference between a liar and a bullshitter is that a liar respects the truth and tries to insert their lies in a calculated way, while a bullshitter talks without any concern for truth.
The problem with society is that people often believe in common practices and follow what others believe, which is rational to get along but not to find the truth.
Science has norms of rejecting authority and skepticism, but day-to-day relies on authority and consensus as a time-saving device.
AI promises to make comprehensive ownership of the full set of facts possible, but there will still be a boundary where we are accepting that the AI is integrating all of the authorities in a way that works.
There is a divide between consciousness and what we call the unconscious, but the term "unconscious" can be misleading.
Wittgenstein criticized the reification of absence, which is similar to the reification of the parts of ourselves that we don't experience as a storehouse of potential conscious mental states.
The mind is an emergent phenomenon, and even with perfect AI and infinite computing resources, we are never going to talk about human-scale experience purely in terms of neurotransmitters and synaptic connections, much less atoms and subatomic particles.
Reductionism and emergence are related issues in discussing the mind, with some believing that the mind can be reduced to more basic facts, while others believe it's a fool's errand and that we'll never talk about human-scale experience purely in terms of microphysical constituents.