ChatGPT can now hold conversations.
Table of Contents
Introduction
OpenAI’s latest iteration of their popular chatbot exhibits Siri- and Alexa-like behavior. It is possible to talk to it and carry on a discussion with it.
OpenAI announced that all ChatGPT Plus subscribers would receive the updated chatbot within the following two weeks.
The San Francisco artificial intelligence startup OpenAI debuted a voice-enabled version of its popular chatbot on Monday. ChatGPT can be spoken to and conversed with in the same way that digital assistants like Alexa and Siri can.
ChatGPT now has the capability to interact with photos for the first time. For instance, if a user uploads a picture of their fridge, the chatbot can suggest meals that can be made from the ingredients they have on hand.
Peter Deng, vice president of consumer and enterprise product at OpenAI, stated, “We’re looking to make ChatGPT easier to use — and more helpful.”
In recent weeks, OpenAI has increased the rollout of its artificial intelligence capabilities. Earlier last month, it released an update to its DALL-E picture generator, integrating the feature into ChatGPT.
After the November launch of ChatGPT, which has now gained hundreds of millions of users, numerous other companies created competing products. OpenAI’s updated bot competes not just with newer technologies like Alexa and Siri, but also with more established ones like Google Bard.
Uncharted Territory
The advent of AI-enabled chatbots has sparked a race to ascertain whether or not this new technology will fundamentally alter the internet economy, rendering current sector leaders obsolete and giving rise to tomorrow’s titans. Take note of these notable robots:
ChatGPT
OpenAI’s ChatGPT, an AI language model, has been making news since last November thanks to its impressive capacity to answer hard questions, compose poetry, develop code, arrange trips, and translate languages. In addition to text, the most recent version of GPT, released in the middle of March, can also respond to photos (and do well on the Uniform Bar Exam).
Bing
Microsoft, OpenAI’s principal investor and collaborator, added a comparable chatbot to its Bing internet search engine two months after ChatGPT’s debut, making it capable of having open-ended text conversations on practically any topic. However, much of the publicity that followed the bot’s release focused on its occasionally incorrect, deceptive, and bizarre responses.
Bard
In March, Google’s chatbot, Bard, was made available to a select group of users in the US and UK. Initially created as a tool for drafting creative pieces like letters and poems, it can now also generate ideas, make blog articles, and answer queries with either factual or subjective responses.
Ernie
In March, Baidu, China’s largest search engine, introduced the country’s first big alternative to ChatGPT. After a “live” demonstration of Ernie, short for Enhanced Representation through Knowledge Integration, was found to be taped, the bot’s premiere was met with widespread disapproval.
OpenAI
Alexa and Siri have been around for a while, offering voice commands for use with mobile devices like phones and computers. Chatbots like ChatGPT and Google Bard, on the other hand, have superior linguistic capabilities and can quickly compose emails, poems, term papers, and riffs on virtually any topic you throw at them.
OpenAI has effectively fused the two forms of interaction. The company prefers voice interactions with its chatbot because it feels more natural to users. ChatGPT boasts five different synthetic speech options, including male and female voices, and the paper claims that they are more believable than those employed by other popular digital assistants.
Customers who pay $20 per month for ChatGPT Plus will begin receiving the updated chatbot within the next two weeks, the business has announced. However, when utilized with iOS and Android devices, the bot only responds by voice.
The bot has a more natural-sounding synthetic voice than many others on the market, although it can still seem robotic at times. However, ChatGPT’s ability to self-correct is one of its many benefits.
While the speech interface of ChatGPT is similar to that of older assistants, the underlying technology is somewhat different. In order to produce language on the fly, ChatGPT is principally powered by a large language model (L.L.M. ), which has been trained on massive volumes of text collected from around the internet.
Outdated Bots
Previous iterations of digital assistants, such as Alexa and Siri, worked more like command centers, with limited functionality; for example, you could only ask them to turn off the lights or ask them about the weather in New York by saying things like, “Alexa, turn off the lights” or “What’s the weather like in New York?” It could take weeks to update the older helpers to support the new instructions. In seconds, ChatGPT can provide an official response to every query you throw at it, however it isn’t always right.
Companies like Amazon and Apple are updating their digital assistants to be more like ChatGPT as OpenAI evolves it to be more like Alexa and Siri.
Amazon showed a new version of Alexa last week that is designed to allow for more natural discussion on “any topic.” The business claims the new L.L.M. and improvements to pace and intonation help it seem more genuine.
Apple has not revealed its strategy to compete with ChatGPT, but two people familiar with the company’s intentions say it has been testing a prototype of its broad language model for future devices.
The new ChatGPT works on desktop computers and mobile devices including the iPhone, iPad, and Android. When shown a picture, chart, or diagram, it can describe it in great detail and answer questions about it. People with vision impairments may find this to be a helpful resource.
Privacy Concerns
In the spring, OpenAI showed off their image tool, but they claimed they wouldn’t release it to the public just yet because of the potential for abuse. They were concerned, among other things, that the program might replace other methods for identifying persons in photographs.
Over the summer, Microsoft released a visual search function in its Bing chatbot that was based on technology developed by OpenAI.
A safety and policy expert at OpenAI has stated that the latest iteration of the bot will now reject all attempts to recognise faces. However, its primary function is to describe other images in great depth. When presented with a Hubble Space Telescope image, for instance, it can provide descriptive paragraphs on the scene captured by the telescope.
The bot can also be used as a teaching aid. The bot can rapidly read and solve a high school-level arithmetic problem complete with words, numbers, and pictures. As a learning tool, or a means of cheating, it has potential.
