Tag Archive : ai bots

/ ai bots

Innovative chatbots

Science fiction has always depicted AI as a kind of machine that can communicate and behave like humans in society. While research on artificial intelligence had already begun by the late 1950s, it wasn’t until the advent of Big Data that AI picked up and became the technological behemoth it is today. 

Artificial intelligence has expanded into countless use-cases and industries, enriching the lives of humans globally. AI started as small computer programs that could play checkers and solve word problems. It has since grown into being able to predict stock markets and election results. While we still don’t have talking robots, or at least not too many, we do have easy access to chatbots. And, some of them have developed into the most strange but interesting use-cases for artificial intelligence the world has ever seen. 

Innovative chatbots

While chatbots may seem human, it’s important to note that they are just lines of code, a string of 1s and 0s floating through cyberspace. However, what they lack in human behavior, they make up for with superior natural language processing, neural networks, and deep learning. This makes chatbots awesome devices that can simulate humans and keep them hooked. Here are some of the most popular ones:

ALICE

It’s hard to talk about chatbots without mentioning ALICE: Artificial Linguistic Internet Computer Entity. It is the first chatbot ever popularized on the web. Developed by Dr. Richard Wallace over 25 years ago, ALICE holds up reasonably well to this day. While her answers don’t always hit the mark, it’s easy to forget how old this code is after even a short conversation with ALICE. 

Sometimes ALICE gives answers that sounded more futuristic than expected at the time. People will remember ALICE for how it spurred the development of numerous, arguably more advanced chatbots over the next couple of decades. ALICE was also the inspiration for the AI computer in the 2013 Academy Award-winning science-fiction film, Her.

Endurance

Designed to help Alzheimer’s patients feel less lonely Endurance is one of the most helpful chatbots for older individuals who suffer from short-term memory loss and dementia. This open-source chatbot was created to identify deviations in conventional branches that could mean problems for immediate recollection, which can be quite the challenge for natural language processing systems. 

Anyone can contribute to the project’s code-base. Though it isn’t fully fleshed out yet, it could potentially provide researchers and scientists more insight into how memory loss works, and how it can be helped. Since the platform is based on the cloud, doctors and family members can review communication logs instantly. This helps identify potential degradation in short-term memory. Further, it takes the burden off family members to constantly be available as a companion.

U-Report

Global child advocacy non-profit organization, United Nations Children’s Fund (UNICEF) is now using chatbots to help citizens in developing countries to speak out about their communities’ urgent needs. The U-Report chatbot enables large-scale data gathering via polls. By regularly sending out polls prepared according to a range of urgent social issues, users can respond with what needs to be tackled first. UNICEF is then able to use this feedback to create or amend potential policy recommendations. 

In Liberia, this rather basic chatbot was able to create major impact waves by finding out whether teachers were convincing students to have sex in exchange for better grades. 86% of the 13,000 Liberian children involved responded about the issues, which led to a collaborative project between UNICEF and Liberia’s Minister of Education to eradicate the problem.

Woebot

Woebot is a chatbot with a cartoon mascot that helps reduce depression through active listening. Psychologist Alison Darcy developed it at Stanford University. It also praises the user with funny GIFs, memes, and encouragement. The New York Times, Wired, and Business Insider has featured this chatbot multiple times over the years.

They praised its, “gentle attentiveness and sensitivity to emotions.” The chatbot is available to those who need a confidence boost, guidance, or even just a friend. Woebot is available for download on  Google Play Store and Apple’s App Store, and users have applauded its funny and likable personality, despite some scripted lines popping up here and there. 

Insomnobot 3000

Created by Casper, which specializes in mattresses and pillows, this is the most well-named chatbot on the list. Casper developed Insomnobot 3000 to help insomniacs find someone to talk to as they lie awake in bed. The chatbot’s website reads, “A friendly, easily distracted bot designed to keep you company when you just can’t fall asleep. Extra chatty between 11 pm-5 am.”The interface is also much more intuitive, as users don’t need to visit a particular website or download an app to talk to the chatbot. Instead, they are required to text the number provided on the website, similar to how you would text a friend late at night. The bot does bring up some product suggestions here and there, but it never becomes overbearing. It was also included on Forbes as one of the most amazing examples of online chatbots in practice. 

DoNotPay

Everyone cannot afford legal counsel, especially when wrongly accused. DoNotPay is a chatbot developed to help users dispute parking tickets, by guiding you through the process without any lawyer interaction. 

If user reviews are to be believed, the service has been quite successful at achieving its goal and empowers users to settle legal matters from the palm of their hands. It’s also helping individuals actively learn about their rights as citizens. While this isn’t the most conversational chatbot, it serves its purpose well and is one of the most strange but interesting use-cases of artificial intelligence. 

Robot Vera

It is a networking Legal and HR chatbot that business enterprises use to solve many issues relating to the recruitment, legal paperwork, etc. smoothly. Robot Vera can improve the workflow and productivity of companies.

Robot Vera assists HR teams by automatically analyzing resume databases and calling candidates that make a good fit. The conversational AI filters out the applications received for a position to merely 10% of the best suitable ones. It then informs candidates about the job description, schedules and conducts telephonic interviews or video chats.

Robot Vera can evaluate the answers of applicants and perform recruitment tasks almost 10 times faster than humans perform. It can hire employees, handle complex office situations, and even fire them if needed.

BabyCentre UK

BabyCentre UK belongs to Johnson & Johnson, a reputed name in childcare. This bot helps answer questions about pregnancy and childbirth. It can calculate the due date of would-be mothers and guide them for preparing for childbirth. It also provides many resources on self-care for moms going through all stages of motherhood.

The bot provides information on how to get pregnant, weekly articles during pregnancy, health care of the mother and the baby, when to call a doctor for your baby, why your baby does not sleep, among many others. It also provides mothers with a forum to share their doubts and experiences.

Watson Assistant

IBM, a leader in the computer manufacturing space, developed this advanced chatbot. It holds the content of varied industries and is pre-trained for industry-specific use cases, using data relevant to that industry. It can understand historical chat, call logs, search queries and also respond to issues using its knowledge base.

The Watson Assistant can be part of your company’s website, messaging channels, customer service tools adapted, and your mobile apps. The chatbot offers a visual dialog editor making your zero experience of coding your power in developing a new feature.

Replika

Replika is a virtual companion chatbot that intends to promote mental wellness. Users can complain or talk about anything, from their failed relationships to their deepest desires. Replika helps people meet their emotional needs. It soothes them when they feel anxious or heartbroken and need an inspiring or comforting chat. Users can select various options for self-motivation, depending on their choice.

Conclusion

As chatbots continue to become an integral part of our society, the trust placed on user privacy and data collection only increases. Today chatbots can send you money, order your pizza, and as shown above, even settle your legal disputes. These can be frightening from the perspective of AI becoming intelligent enough to make humans obsolete. And this is a question developers keep in mind and debate frequently within the AI community. 

AI bots

Whether you’re tracking short-term memory loss or just need a friend, there is a chatbot out there for you. And they continue to learn with every conversation they have.. They’re on Facebook Messenger, Reddit, and even Slack. The last time you opened a chat with a customer service representative, an AI chatbot most likely handled it. If Google Duplex has shown us anything, it’s that AI might become a much more integral part of our lives than previously envisioned. 

Humans have become more comfortable with chatbots and AI. Artificial intelligence systems are learning more about how humans communicate, behave and function. Perhaps we can live in a world where humans and machines live in harmony. After all, most of the time, we just need someone to talk to.

Virtual Assistants - Alexa, Siri, Google Assistant

Siri was introduced as a feature of the iPhone 4S in 2010. While it could only answer simple questions such as “What’s today’s weather like?” and “Who is Barack Obama?”, users praised the potential of the new voice assistant. Quite a feat for that time for a virtual assistant.

Expectations were high, and Siri fell short. Users complained about inaccurate responses to simple questions or commands. If Siri didn’t know the answer to a question, she’d crack a bad joke, which can seem like an unacceptable excuse for not having the ability to answer a question.

While Apple made improvements to its voice assistant, it wasn’t able to meet a lot of high expectations, and that frustrated users.

Alexa

Three years later, Amazon introduced its own voice assistant named Alexa, and it was instantly pitted against Apple’s alternative. Users observed that Alexa was quicker with responses, and was answering more questions right than wrong. Alexa fell short next to Siri when it comes to the fluidity and flow of requests and conversations. Siri could respond to commands better, and it had no problems understanding multiple sentence structures that conveyed the same message.

In 2016, Google came out with an answer to Siri and Alexa in the form of Google Assistant. It became the gold standard for how natural language processing (NLP) should be implemented with a voice assistant. The drawback of Google Home was that it didn’t have the broad integrations that Alexa had with Amazon’s devices.

These three voice assistants are the most popular in the market and each of them has their own strengths and weakness. But, how exactly do they stand against each other? 

The main tests we will conduct for these voice assistants are commands, conversation flow, music requests, home automation, and technology. MKBHD and Undecided with Matt Farell have given us interesting demonstrations and questions that can be used to test each of these three voice assistants. Let’s compare them using the following parameters:

Commands

Voice assistants started off as devices that could answer simple questions such as the time and the weather. Accuracy of response is key here and speed is an additional bonus.

What’s the weather?

Siri, Alexa, and Google Home have no problem answering this. Google tends to have a slight delay in its response generally, but nothing that could test a user’s patience.

How far away is London?

Siri and Google answered this right in miles as the crow flies, while Alexa provided an inaccurate response, or the answer to a different London (there are 29 places in the world called London). 

Conversation Flow 

When humans have conversations, the talking points build naturally and flow from one topic to another seamlessly. For a voice assistant, understanding context while having a conversation is key. 

Conversation

The following questions were asked one after the other to each voice assistant separately.

Who is the 45th President of the United States?

All three voice assistants provide the right answer. Siri cites the source and asks users if they’d like more information.

Where is he from?

When asked immediately after the previous question, Siri and Google fail. Alexa seems to handle context better than its two competitors.

Music

Since all voice assistants communicate with speakers, they need to understand song, artist and album requests. But before we get into their ability to play a track on-demand, its important to note that each voice assistant only plays music from a select set of streaming services. Alexa wins here as it plays from most major services. Google works only with Google Play Music, YouTube Music, Spotify, and Deezer. And Siri, not surprisingly, only plays from Apple Music.

Play Get Lucky by Daft Punk

Simple task. No losers here.

Play the song that goes “like the legend of the phoenix”

Alexa fails here while Siri and Google Assistant get it right.

Home Automation

Home automation refers to command-based control over home appliances such as fans, den lights, television, heaters, etc. Here’s how the voice assistants fared with the following two questions.

Turn off the den lights

All assistants successfully turned the lights off. 

Set the room temperature to 70 F

Google Assistant and Siri got this right, while Alexa adjusted the room temperature to a value between 65 and 70. 

Technology

Siri primarily works on Natural Language Processing (NLP) integrated with Machine Learning (ML), and voice recognition. Alexa operates on similar tech such as Automated Speech Recognition (ASR), and Natural Language Understanding (NLU). The technology isn’t too different from google either, its voice assistant employs NLP and ML.

Yes, the three voice assistants use ML and NLP to understand what the user is saying and to make suggestions or respond to the user’s language input. While the primary technology is the same or at least similar, the end result is what separates the three. As observed in the tasks assigned to them earlier, certain aspects of each voice assistant’s tech, such as the ability to understand speech patterns and words,  give them an advantage and a disadvantage.

Conclusion

The aim isn’t to be diplomatic, but there isn’t exactly a winner among the three. All the voice assistants can, for the most part, do the same things. Alexa has the largest home-integration options among the three, while Google Assistant and Siri are a lot more natural to talk to. 

Virtual Assistants

If you’re big on home automation and having wide music streaming options, Alexa is the voice assistant for you.

If you find yourself comfortable with Google’s streaming services such as Google Music and Youtube, Google Assistant is a smart pick. It also comes with a formidable range of home automation.

And finally, if your household is equipped with Apple’s products, it’s a no brainer to pick Siri, who’s device also has the best speakers among the three. Siri also has an advantage concerning privacy, as it encrypts all data, unlike its competitors that use it for targetted ad campaigns.

As a consumer, your goal is to see which one of these fits your requirement and aligns with what you’re looking for from a voice assistant.