The last decade is already known for an Artificial intelligence (AI) boom and the mass creation of chatbots. While in 2017 Siri, Alexa, or Cortana have become a part of our daily routine, not many know that the chatbots date back to over 50 years. Let’s dive into the history.
Starting from 60s
The first chatbot, Eliza, was created in 1966. It could imitate the answers of a psychotherapist, paraphrasing the questions it was asked.
Parry, the next chatbot, was created five years later. It could imitate paranoid schizophrenic speech. It is interesting that 48% of psychiatrists didn’t detect that they were speaking with a machine. We do not know what the reason is, whether a highly developed chatbot or low-qualified psychiatrists, but it’s a fact.
AI steps in
Scientists created the first AI chatbot in 1988. Jabberwacky was able to imitate human speech in an entertaining way; it remembered everything it was told and then found answers using the context template method.
Siri’s predecessor, named Smarterchild, was created in 2001. It was an important step that proved that people like communication with a smart machine that can give good advice.
Many successful AI bots and apps, like Siri, Google Now, Alexa, Cortana and IBM Watson, came later.
Chatbots in messengers
Nowadays, chatbots are starting to be implemented in messengers. Chatbots are less expensive to develop and are more user-friendly. We think that they will replace mobile apps and social networks in some industries soon.
Telegram was one of the first messengers that implemented and created API for bots. The largest messengers worldwide, FB messenger, WhatsApp, Viber, WeChat, Skype and Kik, created their own chatbot API soon after. Slack, the number one messenger for business communication launched a chatbot store in December 2015.
Photo-editing apps and bots that utilize a neural network and AI to transform images artistically had incredible success in the summer of 2016. For example, the app Prisma was listed in the top 10 apps in the App Store in 77 different countries. One week after being listed in the iOS App Store, Prisma had been downloaded over 7.5 million times. However, as we write below, there is currently a transition from mobile apps to more convenient mobile solutions. People communicate in images and send selfies from mobile phones via messengers.
Icon8 case study
Now it’s time to learn how our team started to create a bot. Even before the creation of Prisma, Icon8 developed an AI neural network chatbot for photo processing. Anyone can load an image, choose one of the six most popular filters and get a great artistic image. The Icon8 chatbot was ranked first on Telegram and third globally on Facebook Messenger and was ranked first among bots by VentureBeat. The usage of the bot was so high that its server usage cost became a problem. The bot received a grant from Telegram CEO Pavel Durov personally to finance the cloud infrastructure.
In addition to an entertainment bot, Icon8 has also developed a business solution based on the same AI technologies. Large marketplaces like Gap, Amazon and many others display images in a particular style: They have a white background, shadows, and a particular crop and size for objects. According to many researchers, this image style increases sales.
Icon8 automates the image processing of shoes and apparel and has reduced the cost of processing one image from 10–100 US cents to a fraction of a penny by using a deep learning network. It will save millions of dollars for one of the biggest e-commerce marketplaces. AI models learn from existing data sets on the client side and get better with every new image.
We believe that similar AI solutions can be scaled in the image-processing industry and used in areas such as video surveillance, self-driving cars and healthcare, among others.
In the next story, you will find out exactly how AI solutions can be implemented in these industries.
Follow us on Twitter, Telegram, Facebook, Reddit and LinkedIn or visit our website to know more about DBrain and the future of Artificial Intelligence.
