The development of interactive artificial intelligence has been slowly progressing over the last 15 years. Even chatbots that are considered advanced by modern standards still can’t hold a candle to the level improv and intuitiveness required for a standard conversation. A recent innovation in artificial text technology may be the key to changing all of that.
Adam King, a Canadian engineer, recently released his extensive artificial intelligence program “Talk to Transformer,” an artificial intelligence language system based on the more extensive system unveiled earlier in 2019 by OpenAI. Talk to Transformer operates along with the same principles of OpenAI’s system, albeit on a smaller and more niche scale.
How It Works
Talk to Transformer works by accepting text input from users. The input can be as long or as short as the users prefer. Talk to Transform then creates additions to this prompted text using an artificial intelligence algorithm designed to predict the flow, direction, and even content of the text.
The program can be used to create just about any type of written content, including stories, poetry, even news articles. Users are encouraged to play around with the generator to test out its predictive levels; the more users who create input for the site, the more the artificial intelligence system can learn. So far, the program can recognize popular characters or media—such as the Harry Potter books and Lord of the Rings films—and even generate dialogue for stories.
Limitations
As with any artificial intelligence program, there are limitations to Talk to Transformer. The current program is not capable of ensuring consistency in the written content. The content that is generated is coherent, at least on a primary surface level, but it does not always abide by structure or order. For instance, generated characters might disappear in-and-out of the generated story for no reason whatsoever. Dialogue is another weak spot for Talk to Transformer: if it creates lines of dialogue, the conversations tend to be unrelated, creating an awkward and strange narrative that doesn’t always make sense.
Progress into the Future
Despite the current limitations of the program, the possibilities for this type of artificial intelligence should not be ignored. Linguistic artificial intelligence may be leading us towards a future when written tasks can be delegated and when artificial intelligence programs can be used in lieu of human interaction for instances such as 24/7 customer service or booking appointments and much more in between.
Anyone can try out Talk to Transformer; simply head to talktotransformer.com and type in one or more prompts to get the program started. Users should take note that the transformer doesn’t always make sense and, because the results are based on an artificial intelligence learning system, the results may sometimes contain questionable content.
Its limitations may be visible, but there is no denying that Talk to Transformer is one notable step towards a revolution in how we create, see, and ultimately use artificial intelligence.
Asif Razzaq is the CEO of Marktechpost, LLC. As a visionary entrepreneur and engineer, Asif is committed to harnessing the potential of Artificial Intelligence for social good. His most recent endeavor is the launch of an Artificial Intelligence Media Platform, Marktechpost, which stands out for its in-depth coverage of machine learning and deep learning news that is both technically sound and easily understandable by a wide audience. The platform boasts of over a million monthly views, illustrating its popularity among audiences.
