Anzeige

What Can’t Chat GPT Do? – Will humans become obsolete?

For months, ChatGPT has been on everyone’s lips, fueled by the media coverage. It seems to be the new revolution when it comes to artificial intelligence, and pretty much everyone is excited behind the scenes about what that means in the future – also for their own business. How will the new AI, for example, change customer chats – which are now often processed by humans when the ready-made answers are no longer sufficient?

And how will social media, but also search engines, change? How can AI help in medicine? And can ChatGPT really answer all questions, or are there also gaps or wrong answers? This is just a very small excerpt of questions that many people rightly ask themselves. Pupils and students are already using ChatGPT to get their homework done. In any case, the OpenAI servers are already heavily frequented around the world around the clock and are often overloaded.

Was kann Chat GPT nicht - Wird der Mensch obsolet

Topic Overview

Anzeige

Of course, we also thought about it and thought about what the new AI that saw the light of day in November 2022 cannot do. Or will humans sooner or later become obsolete in many areas? Because ChatGPT currently works with a database of around 170 billion words and word combinations.

And the next generation of AI should increase this knowledge database by a factor of 100. This would increase the pool of words and word combinations to around 17,000,000,000,000 (17 trillion).

What Can’t Chat GPT Do? – Will humans become obsolete?

For months, ChatGPT has been on everyone’s lips, fueled by the media coverage. It seems to be the new revolution when it comes to artificial intelligence, and pretty much everyone is excited behind the scenes about what that means in the future – also for their own business. How will the new AI, for example, change customer chats – which are now often processed by humans when the ready-made answers are no longer sufficient?

And how will social media, but also search engines, change? How can AI help in medicine? And can ChatGPT really answer all questions, or are there also gaps or wrong answers? This is just a very small excerpt of questions that many people rightly ask themselves. Pupils and students are already using ChatGPT to get their homework done. In any case, the OpenAI servers are already heavily frequented around the world around the clock and are often overloaded.

Was kann Chat GPT nicht - Wird der Mensch obsolet

Topic Overview

Anzeige

Of course, we also thought about it and thought about what the new AI that saw the light of day in November 2022 cannot do. Or will humans sooner or later become obsolete in many areas? Because ChatGPT currently works with a database of around 170 billion words and word combinations.

And the next generation of AI should increase this knowledge database by a factor of 100. This would increase the pool of words and word combinations to around 17,000,000,000,000 (17 trillion).

What Can’t ChatGPT Do?

What Can’t ChatGPT Do?

Understand Emotions:
ChatGPT is an artificial intelligence programmed to understand and respond to natural language. However, it cannot understand emotions as it has no human experience. It cannot interpret emotions or respond in a way that expresses emotions. Although it is able to recognize and respond to emojis, it cannot understand human emotions expressed in text.

Answer questions about current events and topics:
ChatGPT works with a static database that results from the start date of the software. What does it mean that the knowledge database is not old, but it is also not up-to-date from which the answers result. Therefore, it cannot replace search engines whose results are based on continuously updated content around the clock.

Learning from the context:
ChatGPT is a text generation algorithm based on the analysis of large amounts of text. However, it cannot learn from context or evaluate human experiences that have not been provided to it. It cannot understand human experiences or contexts that are not contained in the texts it was trained on. So if a human experience or context not contained in the lyrics is brought into the conversation, ChatGPT cannot respond to it.

Understanding values and morals:
ChatGPT is not able to understand values or morals as it is just an algorithm analyzing texts. It cannot make any moral or ethical decisions because it has no concept of what is “right” or “wrong”. It is programmed only to analyze texts and react to them in certain ways based on the analysis.

Store personal information:
ChatGPT does not store any personal information about users as it is not programmed to do so. It is only programmed to analyze and react to texts. It has no ability to store information or access information outside of the current conversation.

ChatGPT can’t learn on its own either…
…or expand your own knowledge database. ChatGPT can only reproduce what it has been taught. It cannot learn or further educate itself independently. Although ChatGPT has a lot of knowledge, this is only based on what it has been taught beforehand. ChatGPT has no life experience of its own or experience of the world that has been made available to it.

Also, the AI cannot (yet) understand all languages. Although ChatGPT is capable of providing answers in many different languages, there are still many languages it does not understand. But here we can be sure that other languages will be available in the future.

cshow
Ads

ChatGPT cannot solve all human problems
Although the AI has a lot of knowledge, there are many human problems where there is no solution or where it is difficult to find a solution. ChatGPT can only offer the knowledge and skills it possesses and no more than that.

Understand Emotions:
ChatGPT is an artificial intelligence programmed to understand and respond to natural language. However, it cannot understand emotions as it has no human experience. It cannot interpret emotions or respond in a way that expresses emotions. Although it is able to recognize and respond to emojis, it cannot understand human emotions expressed in text.

Answer questions about current events and topics:
ChatGPT works with a static database that results from the start date of the software. What does it mean that the knowledge database is not old, but it is also not up-to-date from which the answers result. Therefore, it cannot replace search engines whose results are based on continuously updated content around the clock.

Learning from the context:
ChatGPT is a text generation algorithm based on the analysis of large amounts of text. However, it cannot learn from context or evaluate human experiences that have not been provided to it. It cannot understand human experiences or contexts that are not contained in the texts it was trained on. So if a human experience or context not contained in the lyrics is brought into the conversation, ChatGPT cannot respond to it.

Understanding values and morals:
ChatGPT is not able to understand values or morals as it is just an algorithm analyzing texts. It cannot make any moral or ethical decisions because it has no concept of what is “right” or “wrong”. It is programmed only to analyze texts and react to them in certain ways based on the analysis.

Store personal information:
ChatGPT does not store any personal information about users as it is not programmed to do so. It is only programmed to analyze and react to texts. It has no ability to store information or access information outside of the current conversation.

ChatGPT can’t learn on its own either…
…or expand your own knowledge database. ChatGPT can only reproduce what it has been taught. It cannot learn or further educate itself independently. Although ChatGPT has a lot of knowledge, this is only based on what it has been taught beforehand. ChatGPT has no life experience of its own or experience of the world that has been made available to it.

Also, the AI cannot (yet) understand all languages. Although ChatGPT is capable of providing answers in many different languages, there are still many languages it does not understand. But here we can be sure that other languages will be available in the future.

cshow
Ads

ChatGPT cannot solve all human problems
Although the AI has a lot of knowledge, there are many human problems where there is no solution or where it is difficult to find a solution. ChatGPT can only offer the knowledge and skills it possesses and no more than that.

Conclusion

Conclusion

Of course, the points mentioned are only a small excerpt of what you should not expect from ChatGPT. Nevertheless, a certain amount of hype is absolutely justified, because from our point of view this is a major advancement in the field of artificial intelligence. The term “intelligence” should be treated with caution, because intelligence is not only based on a large body of knowledge, but above all on a high level of problem-solving skills. And this cannot always be based on factual and logical values, but also includes moral, ethical and human aspects that an AI can never replace.

In any case, it remains exciting to observe the rapid development so far and to see what opportunities will arise for us as humanity if we use them properly. Because even the best technical innovation can and will sooner or later also offer potential for misuse and be used. Here the developers and programmers are asked to take the right steps right from the start. Like any technical revolution, ChatGPT not only offers opportunities, but also other risks. Because the world of work could also change significantly, for example for media professions, and lead to distortions in certain professions.

And the question of responsibilities could also arise in the future. As an example, we assume that not a doctor, but an AI might at some point analyze an MRT image and give recommendations for action. It is then up to the doctor to decide whether to follow the recommendation or not. So who is to be held liable in the event of a wrong decision? If the doctor decides against the recommendation of the AI and something goes wrong, he has to answer questions as to why he did not follow the recommendation. And if the case is the other way around, the AI could give a wrong and momentous recommendation which the doctor follows, and then who is responsible? The doctor, or the programmers of artificial intelligence. This is all still pure speculation, but you have to ask yourself these questions now.

To a large extent, we already see this in the financial world, where computers decide on the approval of loans, or make automated purchases and sales on the stock markets. Questions upon questions that still need to be answered in the future.

Of course, the points mentioned are only a small excerpt of what you should not expect from ChatGPT. Nevertheless, a certain amount of hype is absolutely justified, because from our point of view this is a major advancement in the field of artificial intelligence. The term “intelligence” should be treated with caution, because intelligence is not only based on a large body of knowledge, but above all on a high level of problem-solving skills. And this cannot always be based on factual and logical values, but also includes moral, ethical and human aspects that an AI can never replace.

In any case, it remains exciting to observe the rapid development so far and to see what opportunities will arise for us as humanity if we use them properly. Because even the best technical innovation can and will sooner or later also offer potential for misuse and be used. Here the developers and programmers are asked to take the right steps right from the start. Like any technical revolution, ChatGPT not only offers opportunities, but also other risks. Because the world of work could also change significantly, for example for media professions, and lead to distortions in certain professions.

And the question of responsibilities could also arise in the future. As an example, we assume that not a doctor, but an AI might at some point analyze an MRT image and give recommendations for action. It is then up to the doctor to decide whether to follow the recommendation or not. So who is to be held liable in the event of a wrong decision? If the doctor decides against the recommendation of the AI and something goes wrong, he has to answer questions as to why he did not follow the recommendation. And if the case is the other way around, the AI could give a wrong and momentous recommendation which the doctor follows, and then who is responsible? The doctor, or the programmers of artificial intelligence. This is all still pure speculation, but you have to ask yourself these questions now.

To a large extent, we already see this in the financial world, where computers decide on the approval of loans, or make automated purchases and sales on the stock markets. Questions upon questions that still need to be answered in the future.

Search for other topics:

About the Author:

Michael W. Suhr
Michael W. SuhrDipl. Betriebswirt | Webdesign- und Beratung | Office Training
After 20 years in logistics, I turned my hobby, which has accompanied me since the mid-1980s, into a profession, and have been working as a freelancer in web design, web consulting and Microsoft Office since the beginning of 2015. On the side, I write articles for more digital competence in my blog as far as time allows.

Search by category:

Search for other topics:

About the Author:

Michael W. Suhr
Michael W. SuhrDipl. Betriebswirt | Webdesign- und Beratung | Office Training
After 20 years in logistics, I turned my hobby, which has accompanied me since the mid-1980s, into a profession, and have been working as a freelancer in web design, web consulting and Microsoft Office since the beginning of 2015. On the side, I write articles for more digital competence in my blog as far as time allows.

Search by category:

Popular Posts:

102, 2024

Integrate and use ChatGPT in Excel – is that possible?

February 1st, 2024|Categories: Artificial intelligence, ChatGPT, Microsoft Excel, Microsoft Office, Shorts & Tutorials|Tags: , , , |

ChatGPT is more than just a simple chatbot. Learn how it can revolutionize how you work with Excel by translating formulas, creating VBA macros, and even promising future integration with Office.

2008, 2023

Internet Addiction – A serious look at a growing problem

August 20th, 2023|Categories: Google, Homeoffice, Shorts & Tutorials|Tags: , , |

Internet addiction is just as serious an illness as alcohol or drug addiction. Just that this is a mental illness. In this article we want to go into the phenomenon in more detail and provide assistance.

607, 2023

Main keyboard shortcuts in Windows 10/11

July 6th, 2023|Categories: Homeoffice, Microsoft Office, Shorts & Tutorials, Windows 10/11/12|Tags: |

Entdecken Sie die wichtigsten Shortcuts in Windows 11, um Ihre Produktivität zu steigern. Von allgemeinen Befehlen bis hin zu spezifischen Fenster-Management- und System-Shortcuts, lernen Sie, wie Sie mit diesen Tastenkombinationen effizienter arbeiten können.

107, 2023

Encrypt USB stick – These options are available

July 1st, 2023|Categories: Shorts & Tutorials, Data Protection, Hardware, Homeoffice, Mac OS, Windows 10/11/12|Tags: , , |

Protecting sensitive data is essential. Encrypting USB sticks provides an extra layer of security. Whether it's built-in software, operating system features, third-party software, or hardware encryption, there are numerous options.

Winter Specials 2024: Word & Excel Templates

Special Offers 2023: Word Design CV-Templates

Monthly Technique Bestsellers:

Bestseller 2022-2023 WLAN-Heizkoerperthermostate

SmartHome | Energy & Security

SmartHome | Energy & Security

Bestseller 2022-2023 WLAN-Heizkoerperthermostate
Bestseller 2022-2023 Notebooks

PC & Accessoires

PC & Accessoires

Bestseller 2022-2023 Notebooks
Bestseller 2022-2023 Smartphones

Smartphone & Accessoires

Smartphone & Accessoires

Bestseller 2022-2023 Smartphones

Popular Posts:

102, 2024

Integrate and use ChatGPT in Excel – is that possible?

February 1st, 2024|Categories: Artificial intelligence, ChatGPT, Microsoft Excel, Microsoft Office, Shorts & Tutorials|Tags: , , , |

ChatGPT is more than just a simple chatbot. Learn how it can revolutionize how you work with Excel by translating formulas, creating VBA macros, and even promising future integration with Office.

2008, 2023

Internet Addiction – A serious look at a growing problem

August 20th, 2023|Categories: Google, Homeoffice, Shorts & Tutorials|Tags: , , |

Internet addiction is just as serious an illness as alcohol or drug addiction. Just that this is a mental illness. In this article we want to go into the phenomenon in more detail and provide assistance.

607, 2023

Main keyboard shortcuts in Windows 10/11

July 6th, 2023|Categories: Homeoffice, Microsoft Office, Shorts & Tutorials, Windows 10/11/12|Tags: |

Entdecken Sie die wichtigsten Shortcuts in Windows 11, um Ihre Produktivität zu steigern. Von allgemeinen Befehlen bis hin zu spezifischen Fenster-Management- und System-Shortcuts, lernen Sie, wie Sie mit diesen Tastenkombinationen effizienter arbeiten können.

107, 2023

Encrypt USB stick – These options are available

July 1st, 2023|Categories: Shorts & Tutorials, Data Protection, Hardware, Homeoffice, Mac OS, Windows 10/11/12|Tags: , , |

Protecting sensitive data is essential. Encrypting USB sticks provides an extra layer of security. Whether it's built-in software, operating system features, third-party software, or hardware encryption, there are numerous options.

Winter Specials 2024: Word & Excel Templates

Special Offers 2023: Word Design CV-Templates

Monthly Technique Bestsellers:

Bestseller 2022-2023 WLAN-Heizkoerperthermostate

SmartHome | Energy & Security

SmartHome | Energy & Security

Bestseller 2022-2023 WLAN-Heizkoerperthermostate
Bestseller 2022-2023 Notebooks

PC & Accessoires

PC & Accessoires

Bestseller 2022-2023 Notebooks
Bestseller 2022-2023 Smartphones

Smartphone & Accessoires

Smartphone & Accessoires

Bestseller 2022-2023 Smartphones

Title

Ads

Popular Posts:

Search by category:

Autumn Specials:

Anzeige
Go to Top