OpenAI has officially unveiled its internet-browsing feature for ChatGPT, marking a significant milestone just three weeks after reintroducing the feature in beta following a months-long hiatus.
ChatGPT, the immensely popular generative AI chatbot that has captured the world’s attention over the past year, has been historically limited by its knowledge, stopping at data available up to September 2021. This limitation made it unsuitable as a real-time search engine. However, OpenAI began its journey to incorporate internet services into ChatGPT back in March, a decision accompanied by inherent risks. The live web is not curated in the same way a static training dataset is, which left room for potential misuse by malicious actors and the unpredictability of algorithmic chaos.
In May, OpenAI began integrating web search capabilities via Bing, the search engine owned by OpenAI’s corporate supporter, Microsoft. The functionality was subsequently extended to the ChatGPT mobile app in late June. Unfortunately, the feature was quickly withdrawn when it was discovered that ChatGPT could access paywalled content.
Fast forward to late September, and OpenAI reintroduced the Browse with Bing feature, having refined ChatGPT’s adherence to instructions provided by content owners. Essentially, it now promised to respect the directives in a site’s Robots.txt file, akin to traditional web crawlers.
Now, Browse with Bing is officially accessible to all Plus and Enterprise subscribers, eliminating the need to toggle any beta settings.
In related developments, OpenAI transitioned DALL-E 3 into beta, one month after unveiling the latest iteration of the text-to-image generator.
DALL-E 3 now seamlessly integrates with ChatGPT, alleviating the need for users to meticulously craft text prompts when requesting image creation. ChatGPT takes on the heavy lifting to ensure that the generated image aligns closely with the user’s intentions.
More significantly, with DALL-E 3 seamlessly embedded within ChatGPT, users can receive images as part of their text-based queries without the need to switch between applications.
DALL-E 3 is currently available in beta on both web and mobile platforms. Users can activate the feature by selecting “DALL-E 3 (Beta)” from the GPT-4 tab inside ChatGPT.
This expansion signifies ChatGPT’s evolution beyond being a pure text-based generator, paving the way for audio and visual elements to become an integral part of its capabilities.
In the previous month, OpenAI equipped ChatGPT with the ability to understand and generate speech, allowing users to engage in verbal conversations with the chatbot. This development bridges the worlds of Alexa-style voice assistants with powerful large language models. Users can request ChatGPT to spontaneously create and narrate bedtime stories for their children, although some supervision might be advisable.
Moreover, ChatGPT now empowers users to seek answers using images. Users can upload pictures of objects and discover what they are or find similar items.
Today’s announcement aligns with OpenAI’s overarching goal to transform ChatGPT into a fully integrated, real-time, multimedia generative search engine.
