Programming Bots to Interact Like Humans
Programming Bots to Interact Like Humans
Key Takeaways
- Rule-based chatbots rely on pre-defined conditions and keywords to provide responses, lacking the ability to adapt to context or learn from previous interactions.
- AI chatbots, such as ChatGPT, use large language models trained on massive datasets to simulate human-like conversations and understand conversational context.
- Advancements in AI chatbots include the incorporation of artificial general intelligence (AGI) and physical embodiments, like humanoid robots, showcasing the potential for more interactive and conversational interactions with humans.
MUO VIDEO OF THE DAY
SCROLL TO CONTINUE WITH CONTENT
Chatbots have been a quirky yet useful online tool for some time. The rise of AI-based language models, such as GPT-4 and the ChatGPT chatbot it powers, has given the human-bot-human interface a new flare. But how do AI chatbots simulate human-like conversations? How can a computer simulate conversations with people?
## What Are Chatbots? How Do Chatbots Work?
Before the likes of ChatGPT, Claude, and Google Bard, there were more rudimentary chatbots. These are known as rule-based chatbots or decision-tree chatbots.
A rule-based chatbot doesn’t adapt to situations or understand context and cannot simulate human logic. Rather, they have a series of rules, patterns, and dialog trees set out by the developer that they must adhere to.
Rule-based chatbots follow pre-defined conditions when given a prompt. Keywords are an important factor here. User inputs are scanned by the chatbot for specific words to help it understand what is being asked. Without the ability to understand context, a rule-based chatbot must rely on clues like this to provide a useful response.
Many businesses use rule-based chatbots as a buffer between a customer and a human representative. If you’ve ever tried contacting your energy or cell service provider, you may have been asked to explain your query to a chatbot first. Alternatively, a chatbot may pop up when you visit a website for questions.
Rule-based chatbots can’t answer very convoluted, layered questions. They’re designed to respond to short and simple queries, such as “Change my account details.” A question containing many variables will likely be beyond the scope of a rule-based chatbot, either because it is not trained to interpret natural language or because its database of knowledge is limited.
Rule-based chatbots can’t improve without manual intervention on the development end. This is because they can’t learn from previous interactions.
AI chatbots are also given rules. ChatGPT, for instance, cannot swear or provide criminal advice. However, the way AI chatbots function and interact stretches far beyond what any rule-based chatbot can handle.
How AI Chatbots Work
AI chatbots didn’t start with ChatGPT. Before ChatGPT hit the mainstream, some less advanced chatbots still used AI to interact with their human users.
Take Eviebot , for example. Launched in 2008, Evie uses AI to interact with users. As a learning AI chatbot, Evie can build her conversational skills by noting what other users have typed in the past. In fact, Evie uses the same AI system as Cleverbot, another chatbot that became a mainstream hit in the late 2000s and early 2010s.
But this chatbot is a far cry from the modern versions we use today.
Additionally, Evie isn’t a great informational resource. When we asked Evie how big the sun was, she responded, “Bigger than my future.” While comical, Evie isn’t adept at providing users with facts, regardless of how common they may be. If you’re looking for a more fun-filled or bizarre chatbot experience, Evie may be the right choice for you.
Sites like Cleverbot and Evie are certainly entertaining, but they’re not suited for practical use. In late 2022, the world began to see how incredibly useful AI chatbots could be.
The question remains: how do AI chatbots like ChatGPT simulate accurate conversations with humans? How can they seem almost indistinguishable from a regular person sitting at a keyboard?
In November 2022, OpenAI released a publically accessible version of its GPT-3.5 large language model named ChatGPT. This was the first AI chatbot to showcase an ability to simulate very human-like conversations. We have a dedicated article explaining ChatGPT in-depth , but there are some important pointers to note here.
Firstly, the “GPT” element of the tool’s name stands for “Generative Pre-trained Transformer,” which is a kind of large language model (LLM) . You may have seen both these terms thrown around a lot through 2023, but what do they actually mean?
An LLM is an AI learning model used by all the major AI chatbots you see today. It is powered by an AI algorithm that uses deep learning to operate on an incredibly complex level. All LLMs are trained with very large datasets, giving them a huge reservoir of knowledge to solve issues and respond to queries. ChatGPT-4, for example, was trained with between 1 trillion and 1.7 trillion parameters and terabytes of data (though OpenAI hasn’t revealed exactly how much).
A GPT is a specific type of LLM comprising a neural network capable of deep learning. GPTs are pre-trained models given huge databases of information to learn from. In ChatGPT’s case, this includes text from books, journals, articles, and more. But even with all this data, how does ChatGPT talk to people in a human-like way?
During ChatGPT’s development, it was trained using the reinforcement learning from human feedback (RLHF) method. This form of training uses reinforcement to mold ChatGPT into the desired chatbot. With a reward and feedback model, ChatGPT can understand which responses are useful or “good” and which are not. This method also allows ChatGPT to grasp conversational context better, meaning it can answer prompts more effectively.
ChatGPT’s natural language processing also plays a big role in how it responds to users, including recognizing specific language patterns and sentiments. In its training, the algorithm was provided with examples of human conversations to better understand how humans communicate. The algorithm can even keep note of cues, like greetings and farewells, to monitor the stage of the conversation.
OpenAI has released limited information on GPT-5, the next iteration of its LLM. What’s particularly exciting about GPT-5 (on top of its more up-to-date knowledge base) is that it is rumored to incorporate artificial general intelligence (AGI) into its algorithm. Given that AGI should theoretically be able to simulate human cognition, this may be a game changer.
ChatGPT took the world by storm and continues to do so, but AI chatbots don’t end with OpenAI. Companies worldwide are working to improve their AI chatbots to simulate conversations with people, with some AI chatbots taking things to a physical level.
Take Desdemona, for example, a humanoid robot model that uses AI to communicate.
Created by Hanson Robotics and SingularityNET, Desdemona is the “sister” of the well-known robot Sophia, who has hit many major news headlines for her impressive yet eerie human-like features and temperament.
Unlike Sophia, Desdemona focuses on music and is even part of a band with other human musicians. The AI algorithm draws from a library of preexisting music, allowing Desdemona to sing along to popular songs. The robot has even performed live with her bandmates.
But Desdemona can also talk and hold conversations with people. In 2022, Desdemona was interviewed by YouTube creator Discover Crypto, wherein the creator of her AI algorithm, Ben Goertzel, also answered some questions on AI and its future.
Desdemona’s long-standing joke about keeping humans in aquariums may be unsettling to some, but her ability to respond to non-rehearsed prompts shows the potential AI has to interact with humans in a friendly and conversational manner.
AI Is Only Getting Smarter
Over the past decade, huge strides have been made in the AI field, with chatbots now being able to tell jokes, write essays, translate languages, and provide a huge amount of information. Above all, they have the incredible ability to simulate human conversations. One day, we may see chatbots surpass human ability, but for now, there’s a lot of room for improvement.
MUO VIDEO OF THE DAY
SCROLL TO CONTINUE WITH CONTENT
Chatbots have been a quirky yet useful online tool for some time. The rise of AI-based language models, such as GPT-4 and the ChatGPT chatbot it powers, has given the human-bot-human interface a new flare. But how do AI chatbots simulate human-like conversations? How can a computer simulate conversations with people?
What Are Chatbots? How Do Chatbots Work?
Before the likes of ChatGPT, Claude, and Google Bard, there were more rudimentary chatbots. These are known as rule-based chatbots or decision-tree chatbots.
A rule-based chatbot doesn’t adapt to situations or understand context and cannot simulate human logic. Rather, they have a series of rules, patterns, and dialog trees set out by the developer that they must adhere to.
Rule-based chatbots follow pre-defined conditions when given a prompt. Keywords are an important factor here. User inputs are scanned by the chatbot for specific words to help it understand what is being asked. Without the ability to understand context, a rule-based chatbot must rely on clues like this to provide a useful response.
Many businesses use rule-based chatbots as a buffer between a customer and a human representative. If you’ve ever tried contacting your energy or cell service provider, you may have been asked to explain your query to a chatbot first. Alternatively, a chatbot may pop up when you visit a website for questions.
Rule-based chatbots can’t answer very convoluted, layered questions. They’re designed to respond to short and simple queries, such as “Change my account details.” A question containing many variables will likely be beyond the scope of a rule-based chatbot, either because it is not trained to interpret natural language or because its database of knowledge is limited.
Rule-based chatbots can’t improve without manual intervention on the development end. This is because they can’t learn from previous interactions.
AI chatbots are also given rules. ChatGPT, for instance, cannot swear or provide criminal advice. However, the way AI chatbots function and interact stretches far beyond what any rule-based chatbot can handle.
How AI Chatbots Work
AI chatbots didn’t start with ChatGPT. Before ChatGPT hit the mainstream, some less advanced chatbots still used AI to interact with their human users.
Take Eviebot , for example. Launched in 2008, Evie uses AI to interact with users. As a learning AI chatbot, Evie can build her conversational skills by noting what other users have typed in the past. In fact, Evie uses the same AI system as Cleverbot, another chatbot that became a mainstream hit in the late 2000s and early 2010s.
But this chatbot is a far cry from the modern versions we use today.
As you can see in the screenshot above, Evie isn’t great at answering questions accurately or keeping conversational history in mind. In just a few seconds, the chatbot said its name was Eliza but then changed it to Adam in the next response.
Additionally, Evie isn’t a great informational resource. When we asked Evie how big the sun was, she responded, “Bigger than my future.” While comical, Evie isn’t adept at providing users with facts, regardless of how common they may be. If you’re looking for a more fun-filled or bizarre chatbot experience, Evie may be the right choice for you.
Sites like Cleverbot and Evie are certainly entertaining, but they’re not suited for practical use. In late 2022, the world began to see how incredibly useful AI chatbots could be.
The question remains: how do AI chatbots like ChatGPT simulate accurate conversations with humans? How can they seem almost indistinguishable from a regular person sitting at a keyboard?
In November 2022, OpenAI released a publically accessible version of its GPT-3.5 large language model named ChatGPT. This was the first AI chatbot to showcase an ability to simulate very human-like conversations. We have a dedicated article explaining ChatGPT in-depth , but there are some important pointers to note here.
Firstly, the “GPT” element of the tool’s name stands for “Generative Pre-trained Transformer,” which is a kind of large language model (LLM) . You may have seen both these terms thrown around a lot through 2023, but what do they actually mean?
An LLM is an AI learning model used by all the major AI chatbots you see today. It is powered by an AI algorithm that uses deep learning to operate on an incredibly complex level. All LLMs are trained with very large datasets, giving them a huge reservoir of knowledge to solve issues and respond to queries. ChatGPT-4, for example, was trained with between 1 trillion and 1.7 trillion parameters and terabytes of data (though OpenAI hasn’t revealed exactly how much).
A GPT is a specific type of LLM comprising a neural network capable of deep learning. GPTs are pre-trained models given huge databases of information to learn from. In ChatGPT’s case, this includes text from books, journals, articles, and more. But even with all this data, how does ChatGPT talk to people in a human-like way?
During ChatGPT’s development, it was trained using the reinforcement learning from human feedback (RLHF) method. This form of training uses reinforcement to mold ChatGPT into the desired chatbot. With a reward and feedback model, ChatGPT can understand which responses are useful or “good” and which are not. This method also allows ChatGPT to grasp conversational context better, meaning it can answer prompts more effectively.
ChatGPT’s natural language processing also plays a big role in how it responds to users, including recognizing specific language patterns and sentiments. In its training, the algorithm was provided with examples of human conversations to better understand how humans communicate. The algorithm can even keep note of cues, like greetings and farewells, to monitor the stage of the conversation.
Image Credit: Thanakorn Lappattara/Vecteezy
OpenAI has released limited information on GPT-5, the next iteration of its LLM. What’s particularly exciting about GPT-5 (on top of its more up-to-date knowledge base) is that it is rumored to incorporate artificial general intelligence (AGI) into its algorithm. Given that AGI should theoretically be able to simulate human cognition, this may be a game changer.
ChatGPT took the world by storm and continues to do so, but AI chatbots don’t end with OpenAI. Companies worldwide are working to improve their AI chatbots to simulate conversations with people, with some AI chatbots taking things to a physical level.
Take Desdemona, for example, a humanoid robot model that uses AI to communicate.
Created by Hanson Robotics and SingularityNET, Desdemona is the “sister” of the well-known robot Sophia, who has hit many major news headlines for her impressive yet eerie human-like features and temperament.
Unlike Sophia, Desdemona focuses on music and is even part of a band with other human musicians. The AI algorithm draws from a library of preexisting music, allowing Desdemona to sing along to popular songs. The robot has even performed live with her bandmates.
But Desdemona can also talk and hold conversations with people. In 2022, Desdemona was interviewed by YouTube creator Discover Crypto, wherein the creator of her AI algorithm, Ben Goertzel, also answered some questions on AI and its future.
Desdemona’s long-standing joke about keeping humans in aquariums may be unsettling to some, but her ability to respond to non-rehearsed prompts shows the potential AI has to interact with humans in a friendly and conversational manner.
Over the past decade, huge strides have been made in the AI field, with chatbots now being able to tell jokes, write essays, translate languages, and provide a huge amount of information. Above all, they have the incredible ability to simulate human conversations. One day, we may see chatbots surpass human ability, but for now, there’s a lot of room for improvement.
MUO VIDEO OF THE DAY
SCROLL TO CONTINUE WITH CONTENT
Chatbots have been a quirky yet useful online tool for some time. The rise of AI-based language models, such as GPT-4 and the ChatGPT chatbot it powers, has given the human-bot-human interface a new flare. But how do AI chatbots simulate human-like conversations? How can a computer simulate conversations with people?
## What Are Chatbots? How Do Chatbots Work?Before the likes of ChatGPT, Claude, and Google Bard, there were more rudimentary chatbots. These are known as rule-based chatbots or decision-tree chatbots.
A rule-based chatbot doesn’t adapt to situations or understand context and cannot simulate human logic. Rather, they have a series of rules, patterns, and dialog trees set out by the developer that they must adhere to.
Rule-based chatbots follow pre-defined conditions when given a prompt. Keywords are an important factor here. User inputs are scanned by the chatbot for specific words to help it understand what is being asked. Without the ability to understand context, a rule-based chatbot must rely on clues like this to provide a useful response.
Many businesses use rule-based chatbots as a buffer between a customer and a human representative. If you’ve ever tried contacting your energy or cell service provider, you may have been asked to explain your query to a chatbot first. Alternatively, a chatbot may pop up when you visit a website for questions.
Rule-based chatbots can’t answer very convoluted, layered questions. They’re designed to respond to short and simple queries, such as “Change my account details.” A question containing many variables will likely be beyond the scope of a rule-based chatbot, either because it is not trained to interpret natural language or because its database of knowledge is limited.
Rule-based chatbots can’t improve without manual intervention on the development end. This is because they can’t learn from previous interactions.
AI chatbots are also given rules. ChatGPT, for instance, cannot swear or provide criminal advice. However, the way AI chatbots function and interact stretches far beyond what any rule-based chatbot can handle.
How AI Chatbots Work
AI chatbots didn’t start with ChatGPT. Before ChatGPT hit the mainstream, some less advanced chatbots still used AI to interact with their human users.
Take Eviebot , for example. Launched in 2008, Evie uses AI to interact with users. As a learning AI chatbot, Evie can build her conversational skills by noting what other users have typed in the past. In fact, Evie uses the same AI system as Cleverbot, another chatbot that became a mainstream hit in the late 2000s and early 2010s.
But this chatbot is a far cry from the modern versions we use today.
As you can see in the screenshot above, Evie isn’t great at answering questions accurately or keeping conversational history in mind. In just a few seconds, the chatbot said its name was Eliza but then changed it to Adam in the next response.
Additionally, Evie isn’t a great informational resource. When we asked Evie how big the sun was, she responded, “Bigger than my future.” While comical, Evie isn’t adept at providing users with facts, regardless of how common they may be. If you’re looking for a more fun-filled or bizarre chatbot experience, Evie may be the right choice for you.
Sites like Cleverbot and Evie are certainly entertaining, but they’re not suited for practical use. In late 2022, the world began to see how incredibly useful AI chatbots could be.
The question remains: how do AI chatbots like ChatGPT simulate accurate conversations with humans? How can they seem almost indistinguishable from a regular person sitting at a keyboard?
In November 2022, OpenAI released a publically accessible version of its GPT-3.5 large language model named ChatGPT. This was the first AI chatbot to showcase an ability to simulate very human-like conversations. We have a dedicated article explaining ChatGPT in-depth , but there are some important pointers to note here.
Firstly, the “GPT” element of the tool’s name stands for “Generative Pre-trained Transformer,” which is a kind of large language model (LLM) . You may have seen both these terms thrown around a lot through 2023, but what do they actually mean?
An LLM is an AI learning model used by all the major AI chatbots you see today. It is powered by an AI algorithm that uses deep learning to operate on an incredibly complex level. All LLMs are trained with very large datasets, giving them a huge reservoir of knowledge to solve issues and respond to queries. ChatGPT-4, for example, was trained with between 1 trillion and 1.7 trillion parameters and terabytes of data (though OpenAI hasn’t revealed exactly how much).
A GPT is a specific type of LLM comprising a neural network capable of deep learning. GPTs are pre-trained models given huge databases of information to learn from. In ChatGPT’s case, this includes text from books, journals, articles, and more. But even with all this data, how does ChatGPT talk to people in a human-like way?
During ChatGPT’s development, it was trained using the reinforcement learning from human feedback (RLHF) method. This form of training uses reinforcement to mold ChatGPT into the desired chatbot. With a reward and feedback model, ChatGPT can understand which responses are useful or “good” and which are not. This method also allows ChatGPT to grasp conversational context better, meaning it can answer prompts more effectively.
ChatGPT’s natural language processing also plays a big role in how it responds to users, including recognizing specific language patterns and sentiments. In its training, the algorithm was provided with examples of human conversations to better understand how humans communicate. The algorithm can even keep note of cues, like greetings and farewells, to monitor the stage of the conversation.
How Are AI Chatbots Advancing?
OpenAI has released limited information on GPT-5, the next iteration of its LLM. What’s particularly exciting about GPT-5 (on top of its more up-to-date knowledge base) is that it is rumored to incorporate artificial general intelligence (AGI) into its algorithm. Given that AGI should theoretically be able to simulate human cognition, this may be a game changer.
ChatGPT took the world by storm and continues to do so, but AI chatbots don’t end with OpenAI. Companies worldwide are working to improve their AI chatbots to simulate conversations with people, with some AI chatbots taking things to a physical level.
Take Desdemona, for example, a humanoid robot model that uses AI to communicate.
Created by Hanson Robotics and SingularityNET, Desdemona is the “sister” of the well-known robot Sophia, who has hit many major news headlines for her impressive yet eerie human-like features and temperament.
Unlike Sophia, Desdemona focuses on music and is even part of a band with other human musicians. The AI algorithm draws from a library of preexisting music, allowing Desdemona to sing along to popular songs. The robot has even performed live with her bandmates.
But Desdemona can also talk and hold conversations with people. In 2022, Desdemona was interviewed by YouTube creator Discover Crypto, wherein the creator of her AI algorithm, Ben Goertzel, also answered some questions on AI and its future.
Desdemona’s long-standing joke about keeping humans in aquariums may be unsettling to some, but her ability to respond to non-rehearsed prompts shows the potential AI has to interact with humans in a friendly and conversational manner.
AI Is Only Getting Smarter
Over the past decade, huge strides have been made in the AI field, with chatbots now being able to tell jokes, write essays, translate languages, and provide a huge amount of information. Above all, they have the incredible ability to simulate human conversations. One day, we may see chatbots surpass human ability, but for now, there’s a lot of room for improvement.
MUO VIDEO OF THE DAY
SCROLL TO CONTINUE WITH CONTENT
Chatbots have been a quirky yet useful online tool for some time. The rise of AI-based language models, such as GPT-4 and the ChatGPT chatbot it powers, has given the human-bot-human interface a new flare. But how do AI chatbots simulate human-like conversations? How can a computer simulate conversations with people?
What Are Chatbots? How Do Chatbots Work?
Before the likes of ChatGPT, Claude, and Google Bard, there were more rudimentary chatbots. These are known as rule-based chatbots or decision-tree chatbots.
A rule-based chatbot doesn’t adapt to situations or understand context and cannot simulate human logic. Rather, they have a series of rules, patterns, and dialog trees set out by the developer that they must adhere to.
Rule-based chatbots follow pre-defined conditions when given a prompt. Keywords are an important factor here. User inputs are scanned by the chatbot for specific words to help it understand what is being asked. Without the ability to understand context, a rule-based chatbot must rely on clues like this to provide a useful response.
Many businesses use rule-based chatbots as a buffer between a customer and a human representative. If you’ve ever tried contacting your energy or cell service provider, you may have been asked to explain your query to a chatbot first. Alternatively, a chatbot may pop up when you visit a website for questions.
Rule-based chatbots can’t answer very convoluted, layered questions. They’re designed to respond to short and simple queries, such as “Change my account details.” A question containing many variables will likely be beyond the scope of a rule-based chatbot, either because it is not trained to interpret natural language or because its database of knowledge is limited.
Rule-based chatbots can’t improve without manual intervention on the development end. This is because they can’t learn from previous interactions.
AI chatbots are also given rules. ChatGPT, for instance, cannot swear or provide criminal advice. However, the way AI chatbots function and interact stretches far beyond what any rule-based chatbot can handle.
AI chatbots didn’t start with ChatGPT. Before ChatGPT hit the mainstream, some less advanced chatbots still used AI to interact with their human users.
Take Eviebot , for example. Launched in 2008, Evie uses AI to interact with users. As a learning AI chatbot, Evie can build her conversational skills by noting what other users have typed in the past. In fact, Evie uses the same AI system as Cleverbot, another chatbot that became a mainstream hit in the late 2000s and early 2010s.
But this chatbot is a far cry from the modern versions we use today.
Additionally, Evie isn’t a great informational resource. When we asked Evie how big the sun was, she responded, “Bigger than my future.” While comical, Evie isn’t adept at providing users with facts, regardless of how common they may be. If you’re looking for a more fun-filled or bizarre chatbot experience, Evie may be the right choice for you.
Sites like Cleverbot and Evie are certainly entertaining, but they’re not suited for practical use. In late 2022, the world began to see how incredibly useful AI chatbots could be.
How Do Chatbots Simulate Conversations?
The question remains: how do AI chatbots like ChatGPT simulate accurate conversations with humans? How can they seem almost indistinguishable from a regular person sitting at a keyboard?
In November 2022, OpenAI released a publically accessible version of its GPT-3.5 large language model named ChatGPT. This was the first AI chatbot to showcase an ability to simulate very human-like conversations. We have a dedicated article explaining ChatGPT in-depth , but there are some important pointers to note here.
Firstly, the “GPT” element of the tool’s name stands for “Generative Pre-trained Transformer,” which is a kind of large language model (LLM) . You may have seen both these terms thrown around a lot through 2023, but what do they actually mean?
An LLM is an AI learning model used by all the major AI chatbots you see today. It is powered by an AI algorithm that uses deep learning to operate on an incredibly complex level. All LLMs are trained with very large datasets, giving them a huge reservoir of knowledge to solve issues and respond to queries. ChatGPT-4, for example, was trained with between 1 trillion and 1.7 trillion parameters and terabytes of data (though OpenAI hasn’t revealed exactly how much).
A GPT is a specific type of LLM comprising a neural network capable of deep learning. GPTs are pre-trained models given huge databases of information to learn from. In ChatGPT’s case, this includes text from books, journals, articles, and more. But even with all this data, how does ChatGPT talk to people in a human-like way?
During ChatGPT’s development, it was trained using the reinforcement learning from human feedback (RLHF) method. This form of training uses reinforcement to mold ChatGPT into the desired chatbot. With a reward and feedback model, ChatGPT can understand which responses are useful or “good” and which are not. This method also allows ChatGPT to grasp conversational context better, meaning it can answer prompts more effectively.
ChatGPT’s natural language processing also plays a big role in how it responds to users, including recognizing specific language patterns and sentiments. In its training, the algorithm was provided with examples of human conversations to better understand how humans communicate. The algorithm can even keep note of cues, like greetings and farewells, to monitor the stage of the conversation.
Image Credit: Thanakorn Lappattara/Vecteezy
OpenAI has released limited information on GPT-5, the next iteration of its LLM. What’s particularly exciting about GPT-5 (on top of its more up-to-date knowledge base) is that it is rumored to incorporate artificial general intelligence (AGI) into its algorithm. Given that AGI should theoretically be able to simulate human cognition, this may be a game changer.
ChatGPT took the world by storm and continues to do so, but AI chatbots don’t end with OpenAI. Companies worldwide are working to improve their AI chatbots to simulate conversations with people, with some AI chatbots taking things to a physical level.
Take Desdemona, for example, a humanoid robot model that uses AI to communicate.
Created by Hanson Robotics and SingularityNET, Desdemona is the “sister” of the well-known robot Sophia, who has hit many major news headlines for her impressive yet eerie human-like features and temperament.
Unlike Sophia, Desdemona focuses on music and is even part of a band with other human musicians. The AI algorithm draws from a library of preexisting music, allowing Desdemona to sing along to popular songs. The robot has even performed live with her bandmates.
But Desdemona can also talk and hold conversations with people. In 2022, Desdemona was interviewed by YouTube creator Discover Crypto, wherein the creator of her AI algorithm, Ben Goertzel, also answered some questions on AI and its future.
Desdemona’s long-standing joke about keeping humans in aquariums may be unsettling to some, but her ability to respond to non-rehearsed prompts shows the potential AI has to interact with humans in a friendly and conversational manner.
AI Is Only Getting Smarter
Over the past decade, huge strides have been made in the AI field, with chatbots now being able to tell jokes, write essays, translate languages, and provide a huge amount of information. Above all, they have the incredible ability to simulate human conversations. One day, we may see chatbots surpass human ability, but for now, there’s a lot of room for improvement.
- Title: Programming Bots to Interact Like Humans
- Author: Brian
- Created at : 2024-09-06 23:30:23
- Updated at : 2024-09-07 23:30:23
- Link: https://tech-savvy.techidaily.com/programming-bots-to-interact-like-humans/
- License: This work is licensed under CC BY-NC-SA 4.0.