Sam AltmanCEO of OpenAIenvisions a near future with an Artificial Intelligence (AI) more powerful and capable than ever, as expressed during a recent talk at the Stanford University. ChatGPT-4 “it’s the dumbest model any of you will ever have to use again.” In essence, Altman described ChatGPT as the dumbest AI chatbot, hinting at the future potential of its successors.
He stated that this is because it is “important to ship early and frequently and believe in iterative deployment”, referring to the popular chatbot from OpenAI.
The executive anticipated that the coming model GPT-5 be “really good, materially better” than his predecessor GPT-4which he had previously described as “kind of sucks.”
“One thing we have learned is that me and surprise don’t go well together. People don’t want to be surprised. “People want gradual deployment and the ability to influence these systems,” the CEO said.
In his presentation titled “The Possibilities of AI” (The Possibilities of AI) in Stanford eCorner, Altman He was “optimistic, excited and curious” about the future of AI, just as he felt 19 years ago when he was a student at that same university.
“This is probably the best time to start a business since the appearance of the Internet, given the advances in artificial intelligence,” he stated. The founder of OpenAI He revealed that if he were a student today, he would dedicate himself to AI research in industry, given the large computing capacity required.
Regarding the dangers of this technology, he said that “the greatest risk will not come from a catastrophic event, but from more subtle changes in society,” expressing concern about “the pace at which society will be able to adapt to the great transformations that AI will bring”.
To allow this adaptation, Altman explained that OpenAI seeks to deploy its AI models gradually: “It is important to release early and often, we believe in iterative deployment.” This will allow “that society co-evolves with technology” and provide feedback.
“We don’t know how to productize this in a way that is useful, where the model works really well and where it doesn’t work very well,” he admitted.
About him potential misuse of AI, the CEO said: “With any sufficiently powerful tool. You put the power in the hands of the users of the tool, or you make some decisions that limit what people can do. I think we have a voice in that. I think you all have a voice in that. I think the governments and our elected representatives and democratic processes are the ones that have the most voice.”
He considered that the industry, governments and individuals have a voice in this balance, but “We’re not going to do it perfectly well.”. The technology manager envisions a future with “abundant artificial intelligence available to everyone,” although he estimates that “daily life will not change drastically.”
What it does predict is unprecedented progress: “Every year there will be AI systems that are much more capable than the previous one. AI will get smarter at a rate never seen before”.
When asked about the very high operating cost of ChatGPTwhich is around USD 700,000 diaries, Altman relativized the issue: “Giving people really capable tools and letting them discover how to use them to build the future is a very good and valuable thing, and I am super willing to bet on the ingenuity of you and everyone in the world to discover what to do about”.
“There is probably someone more business-minded than me in OpenAI somewhere worried about how much we are spending, but I don’t care much,” concluded the CEO, thus prioritizing the advancement of AI over the economic implications.
Altman also addressed the peculiar organizational structure of OpenAI, with a non-profit entity that owns a company for commercial purposes. Although he admitted that “it’s not what we would have done if we knew then what we know now,” he defended that they have “a mission that we really care about” and have “adapted the structure as we go.”
Regarding the future of space exploration, the entrepreneur opined that “space is not very hospitable to biological life” and suggested that “if we can send robots, it seems easier.”
Along these lines, he anticipated that in the future access to a certain amount of computing capacity will be considered a human right: “We have arrived at a world where we think of access to a certain amount of computing as a human right, and we have to discover how to distribute that to the people from all over the world”.
In a context of accelerated changes driven by AI, the visionary CEO highlighted the importance of resilience and adaptability: “I believe that in the coming decades, resilience and adaptability will be more important than they have been in a long time.”.
Asked if AI is not would stifle human innovationwas confident that the opposite will happen: “I really deeply believe that people will positively surprise us with better tools. I think the whole story suggests that If you give people more leverage, they do more amazing things.”.
The search for Artificial General Intelligence (AGI) has been the north of OpenAI since its inception. Sam Altman, leader of the company, has reiterated that his “Central vision has long revolved around the search for an AI that can reason as well or better than humans.”
OpenAI’s stated mission is to ensure that this revolutionary technology “benefit all humanity.”
In the race to achieve AGI, questions arise about the tools and resources needed to achieve this milestone. When Altman was asked if the entity that gathers the most chips and computing power will be the winner, his response was cautious but suggestive: “I suspect there will be several different versions.” [de AGI] who are better and worse at different things“, he claimed. “You will have to be above a computation threshold, i guess. But even then, I wouldn’t say I’m sure.”
These statements paint a future in which the AGI will not be a monolithic system, but a diverse ecosystem of specialized models.
Some might excel at specific tasks, while others would offer more general capabilities. What does seem clear is that computational power will be a determining factoralthough Altman acknowledges that even exceeding a certain computational threshold does not guarantee success.
As OpenAI and other organizations venture into this uncharted territory, the path to AGI looks as challenging as it does exciting.
With GPT-5 on the horizon, the CEO and his team are taking leaps and bounds toward that goal. However, the CEO of OpenAI is aware that the race for AGI will not have a single winner, but multiple participants who will help shape the future of artificial intelligence.
During a brief interview with MIT Technology Reviewhe CEO of OpenAIshared his perspective on the role hardware will play in the future of Artificial intelligence (AI).
While acknowledging that current smartphones are extraordinarily capable and that their designers are already incorporating more AI-powered features, some entrepreneurs are betting on a more specialized device for the AI of the future.
Altman mentioned some devices that are already appearing in its orbit, such as the criticized wearable AI Pin of humane, a company in which he is an investor. He is also rumored to be working with former designer Manzana, Jony Iveon some new type of hardware.
“I don’t think it requires new hardware”he said, adding that the type of application envisioned could exist in the cloud.
But he quickly clarified that even if this AI paradigm shift doesn’t require consumers to buy new hardware, “I think they will be happy to have [un nuevo dispositivo]”.
Although the OpenAI entrepreneur finds AI hardware devices exciting, he also hinted that he may not be best suited to take on the challenge himself: “I am very interested in consumer hardware for new technologies. “I am a fan who loves it, but this is very far from my experience”.
- Artificial Intelligence (AI): branch of computer science that seeks to create machines capable of performing tasks that normally require human intelligence, such as learning, reasoning, and problem solving.
- Language model: an AI program designed to process, understand and generate human language in a similar way to how people do.
- GPT (Generative Pre-trained Transformer): A series of language models developed by OpenAI, capable of generating text, answering questions, and performing various linguistic tasks.
- Artificial General Intelligence (AGI): a theoretical field of AI research whose goal is to create software capable of learning and thinking like a human being. AGI is also known as strong AI, and would allow machines to perform a broader range of tasks than weak AI, including creative actions that only humans can do.
- Machine learning (Machine Learning): A subset of AI that allows machines to learn and improve from data and experiences without being explicitly programmed.
- Training data: Information used to train AI models, allowing them to learn patterns and relationships to perform specific tasks.
- Neural network: a computational model inspired by the functioning of the human brain, used in machine learning to process information and recognize patterns.
- Cloud Computing (Cloud Computing) – The practice of using a network of remote servers hosted on the Internet to store, manage, and process data, rather than using a local server or personal computer.
- Wearable: A wearable electronic device, such as a smartwatch or augmented reality glasses, often incorporating tracking and connectivity features.
- Algorithm: a set of step-by-step instructions that describe how to perform a specific task or solve a problem, used in programming and the development of AI systems.
