Sam Altman Just REVEALED The Future Of AI..

So the ability of chatgpt to access the "internet" beyond just the limited supply of web servers, eg your own computer, to look at your files, helps it self train.

Faut taxer les riches

This is one weird dude.

He looks very sad. He also doesn't look very well. He keeps looking away like he wishes he wasn't there. It doesn't look like he has very good well-being. Maybe it's an off day. I can't imagine being practically a celebrity and yet on your shoulders people placing 'world destruction'. Besides strangers calling you a psychopath, like the comment made here.

Anyway. We are all living in a existential terror movie that will never get made, You don't know how long the movie is but you know you'll know when it's over. It's getting palpable now, The malaise, the overarching dread and " what's the point." People I know losing jobs, And no one. I mean no one talks about their future and what they're actually going to do. So sit back, get some popcorn and relax. I feel totally unencumbered by the future! And it feels terrible . 🤮

To boom any industry, you just need to continuously talk about that industry 😅😅

Unless we get wealth consolidation under control, what the future looks like seems pretty clear. Less than 2% of global population will live over 120 years, their children will be genetically improved, their diseases will be addressed, they will enjoy clean food and environment, and they will hold all power including access to Artificial Super Intelligence. Everybody else will suffer fewer and fewer sources of income, and the loss of everything the 1-2% will possess. Same Altman feels comfortable because he’s in that small cadre of the very wealthy. He’s either unbelievably naive, or he fully understands this. Wealth at the expense of others is a mental illness.

The emergence of AI and its necessities

@jamshidi_rahim

I'm thinking these early ai's will quickly pollute the Internet with ai generated content, so you will need to save a snapshot of the Internet the way it was a few years ago to train future models on. The ai generated stuff is garbage compared to dealing with quality human created language and artwork made from the best of the best.

Synthetic data is probably a big reason why humans learn so fast.

That's what imagination and playing are.

We're waiting on it being able to train itself

where is my savage video end that cut right on u talking mid sentence about a video still playing 😀 ?

Can it say the ABC’s yet?

Where can I watch this interview?

TLDR Sam says there is NO sign of plateau in scaling models or diminishing return. But capabilities aren’t evolving at the same speed, models are improving fast in some area but struggle in others. But that’s encouraging.

At a certain point I don’t even know if humans will be good enough to judge the quality of the data. We’re gonna need superhuman AI to know what superhuman quality is.

Summary:

## Sam Altman's Big Reveal: A Deep Dive into OpenAI's Future and the AI Revolution

Sam Altman, the former CEO of OpenAI, recently gave an interview that offered a rare glimpse into the company's inner workings and its vision for the future of artificial intelligence. The interview touched on a range of topics, from the technical details of training AI models to the potential societal impact of advanced AI. Here's a closer look at the key takeaways:

### Data Efficiency: A New Era of AI Training

Altman hinted at a groundbreaking new approach to training AI models, potentially involving synthetic data and a novel architecture. This method, referred to as "SUXSIVA," aims to overcome the limitations of relying solely on real-world data, which can be scarce, expensive, and difficult to obtain in sufficient quantities. This breakthrough, which reportedly played a crucial role in developing GPT-5, could usher in a new era of AI training, enabling faster and more efficient development of powerful models. Altman emphasizes the importance of high-quality data, suggesting that OpenAI is focused on finding ways to learn more from smaller amounts of data, potentially through a combination of synthetic and real-world data.

### The Future of Work: A Looming Social Transformation

Altman acknowledges that AI's rapid advancement will inevitably lead to significant changes in the way we work and interact with the economy. While he doesn't believe in a complete job displacement scenario, he anticipates a profound shift in the nature of work, potentially requiring a reevaluation of the social contract. This shift could necessitate new forms of social safety nets and economic models to ensure a fair and equitable distribution of resources in a future where AI plays a dominant role. Altman suggests that the social contract, currently based on exchanging labor for income, will need to adapt as AI automates more tasks. He points to social safety nets as an example of how society has already begun to adjust to changing economic realities.

### Universal Basic Compute: A Vision for a Post-AGI World

Altman explores the concept of Universal Basic Compute (UBC) as a potential solution for navigating the economic changes brought about by advanced AI. UBC envisions a future where individuals receive a basic allocation of computational resources, potentially more valuable than money, to access and generate various goods and services. This concept, while still in its early stages, offers a glimpse into a potential future where AI empowers individuals with unprecedented capabilities, potentially leading to a more equitable and prosperous society. Altman acknowledges the difficulty in envisioning such a future, as it represents a significant departure from our current economic model.

### GPT-5: A Leap Beyond Current Benchmarks

Altman emphasizes that the next generation of AI models, like GPT-5, will likely surpass current benchmarks, potentially exhibiting capabilities we can't even measure today. He anticipates significant improvements in areas we haven't even considered, requiring us to rethink how we interact with and utilize these systems. This suggests that GPT-5 and future models could be significantly more powerful and versatile than their predecessors, potentially transforming various aspects of our lives. Altman believes that GPT-5 will be "hugely better" in some areas, while surprisingly not as much better in others, a pattern observed with previous model releases. He also acknowledges that GPT-4, despite its capabilities, is still "very, very dumb" compared to what's to come.

### Addressing Controversies: Scarlett Johansson and the OpenAI Board Drama

The interview also addressed recent controversies surrounding OpenAI, including the Scarlett Johansson voice debacle and the drama surrounding Altman's firing. While Altman maintains his perspective on the events, he acknowledges the importance of transparency and addressing concerns raised by the public. Regarding the Scarlett Johansson voice controversy, Altman clarifies that the voice used in the AI model is not intended to be a perfect replica of Johansson's voice, despite the similarities.

Regarding his firing, Altman respectfully disagrees with Helen Toner's account of the events, but acknowledges her genuine concern for a positive AGI outcome. He also clarifies that OpenAI had a gradual release plan for GPT-4, which was already trained before the controversy erupted.

### Interpretability Research: Progress Towards Understanding AI's Inner Workings

Altman acknowledges the importance of interpretability research, which aims to understand the inner workings of AI models. He hints at progress in this area, suggesting that OpenAI may have made breakthroughs in understanding how these models make decisions. This progress could be crucial for ensuring the safe and ethical development of AI, allowing us to better understand and control these powerful systems. Altman acknowledges that OpenAI hasn't completely solved interpretability, but they are making significant progress, potentially suggesting that they have made more progress than previously known.

### A Glimpse into the Future: A World Transformed by AI

The interview provides a glimpse into the future where AI will continue to evolve at a rapid pace, potentially transforming our lives in unimaginable ways. While the challenges are significant, Altman's insights offer a hopeful outlook, suggesting that with careful planning and responsible development, AI can be a force for good in the world.

P.S. by SummaryAdept at POE AI

When society has a problem with AI bad actors it’ll likely be due to SA’s ego

14:46 AIGRID have no comments on lol..

Problem is, Congress is woefully unprepared for the sea change right around the corner. UBI? You'll hear the cries of "Socialism" in Washington. And they'll drag their feet even as the bread lines grow.

synthetic data = youtube videos and hacking devices to use video

As we shift to a mature and intelligent utopian World family, we will be freed from wage slavery, and live with abundance where all people have time to be with their families, create music, art, recreate, go sailing, hiking, cruises, exploring, and enjoying life as it is meant to be just like all of the free abundant creatures of this world. Birds live in paradise and all their needs are met, they do not wage enslave other birds, they share the world, with plenty of room for everyone.

Please make this sam altman take speech classes…that fake low pitch voice he is trying to speak in.. is damn irritating

Ok, that means that GPT5 has been watered down by red teams, previously he said that it would be smarter all across the board and now just that in certain areas. So that means that some of the areas they didn’t know how to handle so they have cut down the model. Well, that is what we could expect.

I’m sure the U.S. military/government/has secretly war gamed how some of this AI will unfold in the public sphere as they’ve likely had it for several years before the public (NSA at least).