Artificial Intelligence (AI) giant OpenAI has launched what it called Model Spec. This is a solution designed to deepen the public conversation about how AI models should behave. Just as it is its culture, OpenAI is also innovating as it looks to make the technology it is pioneering better and safer for all.

OpenAI Model Spec, The Primary Essence

As the company highlighted, its goal is to give humans more control of AI with the agreement of what society deems ethical. It said the Model Spec is a continuation of the ongoing research it continues to sponsor regarding how AI models are shaped.

The unveiled document details how the company’s models are designed to behave during OpenAI API and ChatGPT’s usage. OpenAI clarifies that the model behavior in these AI applications hinges on how they respond to input from humans. These response ranges from tone, personality, and response length.

Over the past few months, users have complained about hallucinations in AI applications, including ChatGPT. These hallucinations are showcased in a way that the application gives a wrong answer to what is required. Sometimes, the AI models act in a lazy way or is outrightly dismissive of users’ prompts.

Series of corrective approaches has been suggested over the years including efforts from Chinese researchers to combat the AI hallucinations. However, with the OpenAI Model Spec innovation, the company is optimistic humans can further join the efforts to make AI better.

Among the key avenues OpenAI noted it will put the Model Spec to use is to support reinforcement learning for AI trainers. It remains unclear whether the Model will be open-sourced for other AI competitors to use. Defining the catch, Sam Altman noted that the Model Spec will improve over time as it will not start operation as a perfect solution.

AI Safety and Regulation

With the technology’s growth rate, fears abound on its potentially harmful impact on humans. This calls for growing push for AI Safety. Despite being a privately held company, OpenAI sits at the forefront of the AI revolution. As a result, it takes center stage in the call for safety across the board.

Other firms in the AI ecosystem including Microsoft and NVIDIA. These firms also remain cautious on the impact of their inventions on public. Most of these tech leaders makes up the government’s AI Safety Board. Based on these, experts believe more concerted efforts will go into making the technology a safe one for all.

Read More: Render Price Prediction: What Awaits RNDR After 50% Weekly Surge

The post OpenAI Teases Model Spec To Control Its Algorithm, But There’s a Catch appeared first on CoinGape.

---

Filed under: News - @ January 1, 1970 12:00 am