Microsoft recently integrated a chatbot into Bing, to disrupt Google’s monopoly in the space. The topmost goal of these chatbots is to save users time. Currently, a Google search result is a never-ending list of articles from which one manually hunts for the answer. The approach is inefficient as most of us don’t even visit Google’s Page 2.
Conversely, when you ask a friend for restaurant suggestions, you don’t expect a detailed commentary about the subject. One expects a crisp recommendation based on the other’s experience. If one wants more detail, they elaborate further. But do chatbots act the same way? Here’s a conversation we had with Google Assistant while writing this article—the technology is far less enthralling.
“Hey Google, find vegan restaurants around me”
“Here’s a list of some of the vegan restaurants in your area”
“Can you pick out the best one”
Repeats the previous answer :(
Currently, it’s trying to remove the drudgery from internet searches and deliver instant, relevant, and accurate answers without the user having to do all the hard work of figuring it out themselves. However, as you can guess, it’s not that easy as each chatbot is a culmination of several large language models.
For instance, Microsoft’s ‘new Bing’ is a result of not just ChatGPT but also Microsoft’s ‘Prometheus’ model for safe, timely and targeted results. Similarly, last year Google took its first initiative with ‘Look and Talk’, an on-device Assistant feature which was an amalgamation of eight different machine learning models.
The ML End Goal (?)
The powerful and dynamic duo of voice assistants and AI is here to stay. Ever since ChatGPT launched, not just tech but every other industry has been obsessed with the idea. To disrupt Google’s search engine monopoly, Microsoft devised the new ‘Bing’ and the rest joined the race. Google retaliated with its $100 billion fiasco ‘Bard’ and the rest jumped on the bandwagon too.
Deedy Das, a former Googler, thinks ChatGPT won’t beat Google search. Google has other language models, such as PaLM and Flan-T5, that are competitive and retraining models is costly. Google search can detect fresh content queries, which ChatGPT can’t. Only 1% of internet power users believe that search is dead and less than 0.1% of Google’s users use ChatGPT.
Even though the numbers are rising, AI-powered search engines and voice assistants have hundreds of unanswered questions and remain risky. Companies are clearly curious to unleash the models’ potential as we see developments unfold routinely. The tech trend seems parasitic but has not managed to convince AI stalwarts. Some experts are astonished while others view it merely as a form of entertainment. Hence, one can conclude, the industry’s “next big thing” remains a dangling subject.
A Paradigm Shift
The ultimate goal of a personal virtual assistant does not seem like a far-fetched sci-fi fantasy any more with ChatGPT taking over the headlines. Ever since movies like Star Trek portrayed interactive computers, sci-fi’s favourite subject has become a billion-dollar solution by the big tech.
The inventor of the world wide web thinks so too. Personal artificial intelligence assistants, such as ChatGPT, are a vision of the future of the web, forecasted Sir Tim Berners-Lee, on CNBC’s Beyond The Valley podcast last week. He said that people can run their personal AI, much like their own version of ‘Alexa’ or ‘Siri, with their data pods. In times to come, Berners-Lee sees each user’s data on their pods which AI could learn from in order to assist.
However, he isn’t the only one forecasting this. At Microsoft Build 2016 event, Microsoft CEO Satya Nadella said that chatbots, as the next big thing, will have “as profound an impact as previous shifts we’ve had”. He further added that the technology has the potential to make computing more accessible. Today, Microsoft has managed to trend on the internet with its latest money-minting launch to run its search engine ‘Bing’ on OpenAI’s next-gen model ChatGPT.
