OpenAI: What is It All About? A Threat to Google, Perplexity, Google, and the AI-Powered Search Engine (GOOGA)
It’s the start of what could become a meaningful threat to Google, which has rushed to bake in AI features across its search engine, fearing that users will flock to competing products that offer the tools first. It also puts OpenAI in more direct competition with the startup Perplexity, which bills itself as an AI “answer” engine. Publishers have accused Perplexity of being directly ripping off their work with its Artificial Intelligence summaries feature.
It explains when to plant tomatoes in another example. You can follow-up questions after the results appear, or click the sidebar to open other relevant links. There is also a feature called visual answers that OpenAI didn’t respond to before publication.
Publishers will have a way to “manage how they appear in OpenAI search features,” the company writes. They can decide not to use their content to train Openai’s models.
Releasing its search engine as a prototype helps OpenAI in a few different ways. It is easier to say it is a prototype if the SearchGPt results are wildly incorrect like when they told us to put glue on our pizza. There’s also potential for getting attributions wrong or maybe wholesale ripping off articles like Perplexity was accused of doing.
This new product has been whispered about for months now, with The Information reporting about its development in February, then Bloomberg reporting more in May. We reported at the same time that OpenAI had been aggressively trying to poach Google employees for a search team. Some X users also noticed a new website OpenAI has been working on that hinted toward the move.
While Openai did not give a demo or interview about the tool, it gave access to unnamed partners and publishers and improved aspects of the search engine based on their feedback.
After OpenAI first introduced ChatGPT in November 2022, early users saw in the chatbot’s ability to dig up and summarize information from the web a potential replacement for conventional web search. The shortcomings of large language models make chatbots imperfect search tools, however. When unsure of an answer, the models will make up facts by drawing on training data that is months or years out of data.
Microsoft’s early efforts with Bing were far from a success, with the AI-powered search engine producing strange, inappropriate, and incorrect answers. Bing’s market share grew only slightly following the overhaul.
Potential copyrighted violations are also a concern. WIRED criticized Perplexity for copying aspects of original journalism, and ignoring requests not to take content from some websites. In OpenAI’s blog post, the company mentions its commitment to publishers: “SearchGPT is designed to help users connect with publishers by prominently citing and linking to them in searches.” News Corp, The Financial Times, and The Atlantic are just a few of the companies that have signed licensing agreements with Openai.