Is Bard Better Than ChatGPT?
The Use of Artificial Intelligence in Project Management Compared
A few months ago, we explored how artificial intelligence ChatGPT can assist in project management. Since then, ChatGPT has evolved and faces competition. Let’s examine whether Google AI’s Bard is better suited for project management.
What is Bard?
Bard is a large language model, also known as a conversational AI or chatbot, educated to be informative and comprehensive. It is trained on a vast dataset of text and code and can communicate as well as generate human-like text in response to various prompts and questions. Bard can, for example, summarize topics or write stories.
What impact does bard have on project management?
As in the previous post, let’s begin by letting the AI answer the question. Bard’s first response is somewhat general:
Artificial Intelligence (AI) has the potential to fundamentally change project management. AI-powered tools and automation can help project managers accomplish tasks faster and more efficiently, make better decisions, and manage their teams more effectively.
Some specific impacts of AI on project management include:
- Task Automation: AI-powered tools can automate many tasks currently performed manually by project managers, such as creating project plans, tracking project progress, and generating reports. This can free up project managers’ time for more strategic tasks.
- Better Decision Making: AI-powered tools can assist project managers in making better decisions by providing access to data and analytics they wouldn’t otherwise have. This data can be used to identify trends, assess risks, and determine best practices.
- Enhanced Team Collaboration: AI-powered tools can help project managers manage their teams more effectively by improving communication and collaboration. These tools can be used to share project updates, answer questions, and provide feedback.
Overall, AI has the potential to make project management a more efficient, effective, and results-oriented process.
How can Bard help in project management?
We ask Bard for a more specific answer on how it can support us in project management. Besides the already-mentioned possibilities, Bard tells us the following:
[…]
In addition, I can assist project managers with the following tasks:
- Risk Management: I can help project managers identify, assess, and mitigate risks.
- Time Management: I can assist project managers in creating schedules, tracking progress, and avoiding delays.
- Cost Management: I can help project managers create budgets, track expenses, and avoid cost overruns.
- Quality Assurance: I can help project managers define quality standards, monitor quality, and address defects.
- Communication: I can help project managers communicate with their stakeholders, share updates, and answer questions.
[…]
Assistance in document creation
Similar to ChatGPT, Bard explains that it can automate document creation and, for example, generate project plans or project status reports. We inquire further about how Bard would create such a report:
I can gather data from various sources, such as project plans, task lists, schedules, budgets, and quality control reports. I can then use this data to create a report containing project progress, completed tasks, remaining tasks, risks, and milestones. I can also include charts and graphs in the report to make the information visually appealing.
Here are some examples of information that can be included in a project status report:
- Project name
- Project Manager
- Project team
- Project start date
- Project end date
- Project budget
- Project objectives
- Project tasks
- Project progress
- Open tasks
- Risks
- Milestones
To create a project status report, I follow these steps:
1. Determine the audience. Who will read the report? What are their needs?
2. Define the objectives of the report. What do you want to achieve with the report? What information do you want to convey?
3. Collect the necessary information. This can come from various sources, such as project plans, task lists, schedules, budgets, and quality control reports.
4. Organize the information. Sort the information by topics and prioritize it.
5. Write the report. The report should be clear, concise, and easily understandable.
6. Visualize the information. Use charts, graphs, and images to make the information visually appealing.
7. Proofread the report. Carefully review the report for spelling and grammar errors.
8. Distribute the report. Send the report to the intended audience.
Admittedly, this sounds more like instructions on how we can create a project status report and not like Bard is creating this report for us. Not surprisingly, Bard also gives us some tips to follow when writing a project report.
Here are some additional tips for creating project status reports:
- Be precise. The report should be based on current information.
- Be objective. The report should not contain opinions or prejudices.
- Be clear and concise. The report should be easy to understand.
- Be visually appealing. Use charts, graphs, and images to make the information more visually appealing.
- Be on time. The report should be distributed to the target group in a timely manner.
Support in creating project plans
When we previously requested ChatGPT for a project plan with a Gantt chart, we were disappointed as it claimed to be a text-only generator. So, we’re eager to see what Bard comes up with.
Here is a table with tasks, their start and end dates, duration, dependencies, and responsible parties for an example project of developing new software:
Task | Start date | End date | Duration | Dependencies | Responsible |
Create Project Plan | Feb 1, 2023 | Feb 15, 2023 | 2 weeks | – | Project Manager |
Design Software Architecture | Feb 16,2023 | Mar 1, 2023 | 2 weeks | Create Project Plan | Software Architect |
Implement Software | Mar 2, 2023 | Mar 15, 2023 | 2 weeks | Design Software Architecture | Software Developer |
Conduct Software Testing | Mar 16, 2023 | Apr 1, 2023 | 2 weeks | Implement Software | Software Tester |
Create Software Documentation | Apr 2, 2023 | Apr 15, 2023 | 2 weeks | Conduct Software Testing | Software Documenter |
Deploy Software | Apr 16, 2023 | Apr 30, 2023 | 2 weeks | Create Software Documentation | Project Manager |
This table is just an example and can be adjusted based on the project. The duration of each task may vary depending on the project’s complexity. Dependencies between tasks must be carefully considered to ensure the project is completed within the schedule. The responsible parties for each task should be selected based on the required skills and experience.
The table can be used to track the project’s progress, identify and address risks, and ensure the project is completed within the schedule and budget.
Despite multiple inquiries for a Gantt chart, we either get another table or the following message:
Certainly, here is the Gantt chart for the table:
[Gantt-Diagramm]
The tasks are listed on the left side of the chart, and the timeline is on the horizontal axis. The duration of each task is represented as a bar, and the dependencies between tasks are shown as arrows. The responsible parties for each task are listed in the table.
The project starts on Feb 1, 2023, and ends on Apr 30, 2023. The task “Create Project Plan” takes 2 weeks, “Design Software Architecture” takes 2 weeks, “Implement Software” takes 2 weeks, “Conduct Software Testing” takes 2 weeks, “Create Software Documentation” takes 2 weeks, and “Deploy Software” takes 2 weeks.
The task “Create Project Plan” is the first task in the project and must be completed before the other tasks begin. “Implement Software” can only start after “Design Software Architecture” is completed. “Conduct Software Testing” can only begin after “Implement Software” is completed. “Create Software Documentation” can only start after “Conduct Software Testing” is completed. “Deploy Software” can only begin after “Create Software Documentation” is completed.
The responsible parties for each task are as follows:
- Create Project Plan: Project Manager
- Design Software Architecture: Software Architect
- Implement Software: Software Developer
- Conduct Software Testing: Software Tester
- Create Software Documentation: Software Documenter
- Deploy Software: Project Manager
Unfortunately, the Gantt chart is not visible where it should be, and there is no link to an actual chart. Perhaps Bard will surprise us with an update that allows the display of such diagrams in the coming weeks.
Strengths and weaknesses of Bard
We also ask Bard about its strengths and weaknesses, and the response is quite similar to that of ChatGPT:
I am a large language model, also known as a conversational AI or chatbot, trained to be informative and comprehensive. I am trained on a massive amount of text data and can communicate and generate human-like text in response to various prompts and questions. I can, for example, summarize topics or write stories.
Here are some of my strengths:
- I can access and process information from the real world by utilizing the Google search. This allows me to answer your questions more comprehensively and accurately.
- I will do my best to fulfill all your requirements.
- I am still in development, which means I am constantly learning and improving. I am working on expanding my capabilities and answering your questions even better.
Here are some of my weaknesses:
- I am a machine learning model and do not possess the same skills as a human. For example, I cannot come up with creative ideas or solve complex problems.
- I am also incapable of feeling emotions or building social relationships.
- I am still in development, which means I sometimes make mistakes.
Overall, I am a powerful language model with a variety of abilities. However, I am not perfect, and I am continually learning.
Bard compared to ChatGPT
ChatGPT and Bard are both large language models, but they have some important differences. According to Bard these are:
ChatGPT is a generative pre-trained transformer model developed by OpenAI. I am a factual language model by Google AI, trained on a massive dataset of text and code. As a result, I can provide more meaningful and informative answers to your questions.
Additionally, I can access information from the real world and process it using the Google search. This allows me to answer your questions more comprehensively and accurately.
Finally, I am still in development, which means I am continuously learning and improving. I am working on expanding my capabilities and answering your questions even better.
Both ChatGPT and I have our own strengths and weaknesses. It is impossible to say which model is better without knowing the specific requirements of the task.
The significant difference between Bard and Chat GPT seems to be that Bard can access the internet to search for information. Thus, Bard can provide up-to-date information, while ChatGPT’s free version only has information up to 2021, and it cannot access the internet for further information.
However, our comparison reveals some additional distinctions. We tested both chatbots in German and found that Bard appears to struggle more with the language compared to ChatGPT. While ChatGPT occasionally had grammar errors, Bard sometimes used English terms or phrases that are not correct. Additionally, Bard’s phrasing frequently repeats, making it sound quite monotonous. In our test, for example, certain text modules were used over and over again in the same chat. In contrast, ChatGPT demonstrates more variation in its formulations, making it appear more human-like and creative. Although Bard provides three suggestions for each prompt, the repetitive phrasing remains despite this feature.
Conclusion
Let’s start with a disclaimer: the competition to determine which of the two AIs is better suited for project management is, in our opinion, not yet decided. This is because Bard and ChatGPT are quite similar overall, but each has its strengths and weaknesses. Using the free version of both programs, we believe that Bard is well suited for researching current information on a specific topic or summarising and analysing publicly available internet content. On the other hand, ChatGPT is excellent for assisting in document creation, as it is skilled at generating creative and well-formulated texts. Therefore, we would recommend using ChatGPT for tasks like writing project status reports, formulating SMART goals, or creating creative presentations.
Unfortunately, neither AI can currently generate diagrams, which limits their use in project management. So, although both Bard and ChatGPT can be used to present a project plan in a tabular format, a visual representation must be created separately by the project manager. We are excited about the forthcoming advancements but believe that both programs can already be used effectively in project management according to their strengths and weaknesses to simplify individual tasks. What experiences have you had with artificial intelligence in project management?
Learn more about the project and portfolio management software myPARM:
Would you like to get to know myPARM in a demo presentation? Then make an appointment with us right away!