OpenAI is not the only winner in the AI ​​arms race

AI For Business

OpenAI CEO Sam Altman has become the de facto face of the generative AI boom. However, his company faces many challenges.
Kevin Deitch/Getty Images

  • Last year, OpenAI’s ChatGPT became a big hit and changed the way people thought about the future of work.
  • But while OpenAI has become synonymous with AI, the race hasn’t stopped.
  • Privacy concerns and data protection drive many OpenAI customers to more niche competitors.

The good news for OpenAI is that it has essentially become synonymous with the current generative AI boom.

The company’s overnight success, ChatGPT, has emerged as the gold standard for AI chatbots, forcing it to catch up to its larger competitor, Google. OpenAI has maintained its momentum by rapidly improving ChatGPT, both by making it smarter and by signing deals with Microsoft and others to bring it to more places.

The bad news is that success can be a double-edged sword. OpenAI’s first-mover advantage is unlikely to dominate the space forever due to privacy rules, regulations, and good old market competition, as competitors like Google and international players like Baidu make their move. is already declining rapidly, leaving huge opportunities for

The crux of the matter is that ChatGPT itself is what is called a general purpose chatbot platform. It is designed to be as useful as possible for a wide range of users across industries and lifestyles.

Experts and industry insiders believe that in the real world there are real benefits to building AI models that are very good for their specific purposes and built to meet the stringent requirements of international privacy and safety regulations. says there is

“I don’t think there is one master AI model that everyone uses,” said Naveen Rao, co-founder of AI company MosaicML.

So while OpenAI’s position as a pioneer in the field is well established, the path to becoming the next great tech giant is less clear.

OpenAI faces real challenges

Since the release of ChatGPT, OpenAI has faced criticism, schools have banned its use to prevent cheating, and regulators in countries such as Italy have banned OpenAI entirely due to privacy concerns. Prohibited.

Other AI modelers see the concerns around OpenAI as a growing opportunity and want to jump at the opportunity.

Meanwhile, leaders including Elon Musk are calling on the industry to slow down the development of AI. The fact that many are unwilling to do so for fear of losing market share proves that OpenAI faces stiff competition who are not waiting to see what happens next.

But while OpenAI is poised to become a giant in this space, it will face stiff competition of its own.

ChatGPT’s Ultimate Weakness: Not Niche Enough

ChatGPT’s big selling point is that it can answer almost any question. It reads large amounts of data, mostly collected from the broader Internet, to find the correct patterns to generate answers to queries.

Not all industries believe their data should be part of the ChatGPT grand public experiment. This belief was strengthened after a bug leaked ChatGPT’s conversation history. Companies such as Walmart, Amazon, and even partner Microsoft warned employees against entering sensitive information on ChatGPT. Amazon, without missing a hype, directed programmers to use his own AI model called CodeWhisperer.

OpenAI’s determination to dominate the field means that it has created more general AI models that others can add to. In that regard, OpenAI licenses his ChatGPT to other organizations. This allows startups and other big brands such as partner and investor Microsoft to build on its blockbuster products.

However, it also leaves the door open to a market for more specific AI models, flocked by organizations in sensitive industries.

Rao says healthcare and financial services companies want to use powerful AI, but they also want to keep data securely and train the AI ​​only on relevant information.

AI is not winner-take-all

Competition in AI means more choice, which can foster better innovation, experts say.

Kari Briski, vice president of AI software at NVIDIA, said: “For large language models, one size model does not fit all scenarios.”

Briski says the ability to customize to your specific needs makes large language models even more valuable to your business. And for AI to thrive further, we can’t create one-size-fits-all.

OpenAI may have moved its AI to the masses in the first place, but its very business model to date has meant that eventually other companies will take advantage of it and serve what it can’t. To do.

Source link

Leave a Reply

Your email address will not be published. Required fields are marked *