In his latest venture, Altman has created an interface which is able to allow for human-computer interaction in a way that is more sophisticated than any other installed currently. The interface, ChatGPT, uses large language models in order to create a more natural and organic user experience. Despite the fact that ChatGPT is still relatively new, Altman sees it as a stepping stone towards even more advanced interfaces which will allow humans and machines to communicate in ways that are both fluently and intuitively.
Altman went on to say that we are reaching a point where it may be more effective and efficient to invest in smaller models that can adequately cover a specific area of law. This is partially due to the fact that larger models often end up becoming expensive and difficult to manage, not to mention they can become overcrowded which can lead to inaccuracy.
With the advent of powerful artificial intelligence and big data, size is becoming less and less important in determining a model’s quality. Chip speed races used to be important because they demonstrated the power of a particular piece of technology, but now that AI and big data are more prevalent, size is not as important. Though parameter count will surely continue to increase, Altman believes that we should focus on other factors when assessing the quality of a model.
Apple executive Craig Federighi says that the company’s focus is on rapidly increasing capability, even if that means decreasing parameter count over time. This philosophy likely applies to iOS 11 as well, which introduces a number of new features but doesn’t feature any major changes to the core interface. iOS 11 also upped the ante with performance by introducing features like support for Animojis and Super Retina Display resolutions.
Altman is often admired for his ability to make large bets on innovative technologies, and then move quickly to turn them into successful businesses. OpenAI is no exception – the company has made a number of large bets in artificial intelligence (AI), with the hope of creating smarter and more autonomous machines. The most high-profile example of this was OpenAI’s creation of AlphaGo, an AI system that beat one of the world’s top go players at their own game – a testament to Altman’s willingness to take risks and invest in early-stage technologies.
While other businesses have failed due to unsuccessful efforts in developing a successful product, Persistent has been able to successfully create and market their product for seven years. By grindsweating every detail and utilizing a strong focus on customer service, the company was able to build a thriving business. While some businesses may not be willing or able to take the time needed to develop a successful product, Persistent has accomplished this by focusing on long-term goals and maintaining continuous hard work.
Many people feel that OpenAI’s approach to artificial intelligence is too reckless, and that the company should pause for six months to contemplate their methods. Bob Izzo, one of the founders of OpenAI, defended their practices in a letter addressed to tech leaders around the world. In it he sets out several points outlining why they believe their approach is correct and why they believe that pausing would be a hindrance rather than a help.
My understanding is that NASA is taking the time to study the safety model for their newly released spaceplane, GPT-4. This is welcome news, given the numerous issues with GPT-3. The Trump administration has been pushing for increased national exploration and spaceflight, and it would seem that NASA is taking these concerns seriously.
Some argue that portions of the letter were misguided and could have done more harm than good. For example, while the letter urged representatives to continue to voice their concerns and push for change, some feel that it reminiscences of dictatorial or oppressive governments. Additionally, many felt that its focus on Syrians as a
Some members of the Artificial General Intelligence community are concerned that the safety bar has to be raised, as artificial intelligence capabilities become more and more serious. However, Dr. Michael Nielsen disagrees, and believes that additional measures should be taken to ensure safety is maintained before increasing AI capabilities any further. He thinks that caution and a focus on safety are essential when advancing AI technology, and suggests some possible ways in which this could be accomplished.
Altman’s openness about the safety issues and limitations of the current model is commendable. He recognizes that sometimes he and other company representatives say “dumb stuff,” but he’s willing to take that risk because it’s important to have a dialogue about this technology. If more people were open about these issues, hopefully manufacturers could develop safer models in the future.
OpenAI is trying to get people to think about and adapt their institutions to deal with the potential implications of AI. Their goal is to create a world where everyone is actively engaged in making sure that this technology progresses in a positive way.