Fashion’s Dreams for ChatGPT | BoF

Everyone is excited about ChatGPT.

Conversational artificial intelligence is eerily good at answering questions with prompt responses and compelling human text replies. Unlike new technologies that seem to sit on the sidelines for years while companies look for ways to leverage them, they are rapidly attracting companies looking to put their capabilities to practical use.

Startups are trying it out for simple administrative tasks. Fanatics, which sells sports memorabilia, sees this as a way to power up its customer service chatbot. Mint Mobile, a telecommunications company owned by actor Ryan Reynolds, made a copy of a recent commercial.

Microsoft now plans to integrate AI models from ChatGPT developer OpenAI into its consumer and enterprise products, announcing a “multi-year, multi-billion dollar investment” in the company. bottom. The investment also produced the DALL-E image generator, already touted as disruptive technology. for fashion design.

Google seems concerned. ChatGPT’s potential to transform online search comes after the tech giant reportedly called founder Sergey Brin and his Larry Page (both of whom stepped away from his day job in 2019) to help the company reportedly prompted him to consider his AI strategy.

It’s too early to predict what the exact results will be for fashion. You may be able to improve your method.

ChatGPT is a type of AI called a large language model. They are trained on vast amounts of data and, in the case of ChatGPT, fine-tuned by human supervisors. Noteworthy is the ability to perform a series of different tasks. Currently, the data on which models like ChatGPT are trained tends to be publicly available online, but the greatest potential for business lies in the fact that companies integrate tools with their own data. and started writing more specific applications.

One area of ​​fashion where this type of AI can be imagined to outperform the status quo is online customer service. Chatbots have been around for a while, but they aren’t always effective. Asking a question may display a list of related topics and links to entries on the FAQ page. AI-powered chatbots like ChatGPT may actually be able to tell you the status of an order rather than telling you how to check it yourself.

In theory, you can even go one step further with proper training. If a retailer collected data on why people are returning items, and used AI to analyze it, and analyzed other sources like user reviews, a chatbot could provide an idea of ​​how the item would fit. We may provide guidance on what to wear and offer suggestions on sizing.

Another use is to create copy for everything from marketing to product pages. ChatGPT is definitely a great tool for creating first drafts and human refinement. However, as the tools become more powerful, they could also be used to create micro-targeted ads or create personalized copy for each individual customer (assuming the site can recognize them). Data privacy measures make it harder to track shoppers, so it may only work for customers who are logged into their account.)

Beyond the retailer’s own operations, it may also be affected. As conversational AI begins to change the way people search online, it could change the way shoppers discover new products. For example, instead of searching for “best running sneakers of 2023” and seeing lists compiled by various sites, customers expect search engines to digest all the information out there and provide a direct answer. You may just hope.

These types of changes may still be a long way off. The hype about new innovations tends to get ahead of reality, and innovations don’t always follow the path everyone expects. AI has been touted as having the potential to take over jobs from driving a car to packing boxes, but full automation has always been out of reach. The cost of developing large language models is also high.

Technology also has its limits. ChatGPT replies may look authoritative, but they may contain inaccurate or meaningless information. Sam Altman, CEO of OpenAI, said: Tweeted in December“It would be a mistake to rely on it for anything important now. This is a preview of progress. There is a lot to be done in terms of robustness and veracity.”

One app that simulates chatting with historical figures has come under fire because historical figures sometimes tell blatant lies. No retailer wants to make the same mistake of chatbots providing misinformation or, worse, alienating shoppers with offensive language or remarks.

Large companies and investors rushing into generative AI hope that these flaws will eventually be fixed. There is now a race between OpenAI and its competitors to transform these AI models from fun toys into essential tools used by mainstream businesses and consumers.

Source link

Leave a Reply