- 30 May 2024
- 5 Minutes to read
- DarkLight
A Paperless Office is not yet a reality - with or without AI
- Updated on 30 May 2024
- 5 Minutes to read
- DarkLight
![](http://substackcdn.com/image/fetch/w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff13c31e5-7dc8-41a3-b09c-98a6da9db7a2_897x424.jpeg)
I am pretty open-minded and read a lot about automation tools and AI.
I remember when Robotic Process Automation (RPA) tools came out, and it sounded like they were going to be the end-all for automating and eliminating tedious tasks. I know RPA has helped many companies, but when it comes to office and administrative tasks, I haven’t seen much headway.
For one thing, the RPA tools (at least the ones I tried) are complicated. You can’t just hand over the application to an administrative person, have them build the logic, and start using the tool. Secondly, until a company and/or government agency has every document in electronic form, a document will have to be scanned.
I can’t picture a robot walking around scanning documents or offices with any type of robotics in them. Most offices are cramped for space and are doing their best to work within their environment. (Especially government agencies.)
Now we have AI. From all the recent articles and news, it sounds like you can bring up your "CoPilot or bot,” type in a question about what you need, and you are good to go. It’s like the “Easy” button that will do everything for you.
Yes, many people are using these tools, doing good things, and increasing their productivity.
But in the world of paper and document processes, there is no Easy button yet that I have found.
Scanning and document capture
There is still a lot of scanning being done, and new scanners are purchased every day. I tried to visualize a robot, with a scanner in its chest, walking around and gobbling up the paper. That would be pretty cool. No boxes, paper cuts or storage costs.
Many companies have not invested in capture tools. They buy a scanner, and their employees get basic training on the software included with the scanner. Then, they start scanning. You see documents scanned upside down, the back side of pages missed, and poor-quality images.
Document capture tools like OCR, classification, and automatic indexing have improved greatly. The quality of the images has improved immensely, and tedious data entry has decreased substantially. These tools aren’t part of the AI craze. They came out many years ago and have served office staff very well.
The technology is there to automate and increase productivity. The caveat is that for these tools to do their job, the human factor has to do theirs. The tools aren’t free, some are quite costly. Users have to be trained, and the processes have to be documented step-by-step so that the full value of the investment is received.
Lessons Learned
I operated a service bureau for many years and have also set up and implemented many scanning departments for customers.
The success of a scanning operation comes down to the scanner and software purchased, management and training.
Successful outcomes have occurred with these conditions:
The company/agency worked with a knowledgeable vendor to select the right equipment and software for their environment. The number of pages to be scanned daily, the paper size, the quality of documents, and the final storage of documents play big parts in selection.
End users were brought into the process and purchase discussions. It is so important to listen to the people who do the work, day in and day out, and document their processes. Then, select and design a solution that fits the unique needs of each area.
PDF files on a shared server are how many companies/agencies manage their documents. That’s okay if they have standard filing naming conventions and can control where files are saved and accessed.
Enterprise document management/records management solutions offer many components and options for scanning, indexing, classification, and filing. They can be costly and require a lot of time and support from IT departments, so it is important to devote the time and resources when considering an enterprise solution.
Can AI help?
Scanning and Document Prep
I don’t see how AI can help with scanning or document prep. Someone/something has to get the documents/files for scanning, remove staples and prep for scanning, and put the pieces of paper in the scanner. I can’t imagine companies or government agencies replacing people with robots to do these tasks. Until no one uses paper anymore, I don’t see this job being replaced.
Indexing and Data Entry
As mentioned, there are many tools to help with automatic indexing and data entry, and these came out well before the current AI craze.
AI tools will most likely be available that can take over the human side of document management, such as assigning metadata or indexing and saving documents correctly for future retrieval. Maybe AI will be able to ensure 100% accuracy on documents that need OCR or ICR for full-text search.
Search
Search can still be an issue for many because of poor indexing, metadata, file naming, and file-saving practices.
I tried ChatGPT's document summary tool and the Adobe Acrobat AI Assistant. The main value I saw with these is for marketing. The tool's ability to summarize the content so that you don’t have to read each article definitely saves time.
Marketing literature is normally not the same as an HR dispute, contract issue, or other situation in which humans need to read through documents and make decisions. Many court cases have been won or lost based on the document discovery of the two parties.
The other downside I saw was only being able to do one document at a time, versus a folder or set of documents. Also, the possibility that the documents ran through the summary tool are retained for training data.
This is a screenshot of the Acrobat AI Assistant. I used it on a document retention document. It is similar to ChatGPT’s summary tool.
![](http://substackcdn.com/image/fetch/w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F16f12228-626e-4033-a268-80c1b6cb221b_1470x550.png)
Here is a link to an article in ZDNet - https://www.zdnet.com/article/how-to-use-chatgpt-to-summarize-a-book-article-or-research-paper/
“How to use ChatGPT to summarize a book, article, or research paper”
“Copying and pasting a URL into ChatGPT won't yield you the best results. If there is a specific paragraph or sentence you're struggling with, you can copy and paste it into ChatGPT and ask it to explain it to you. Still, it's best to read the PDF and use the chatbot as a summary tool and not as an educator. “
In Summary
We have come a long way with the use of software to create fillable forms, populate forms with data, compose documents, and distribute documents. But there is still a lot of paper flowing around and printers being used.
I am sure there are many who are hopeful that one day, they can scan or import a document and not have to index it, classify it, or save it anywhere specifically. And when they need to find it, they will just bring up their “assistant” and type something in and it will miraculously bring up the exact document they want. I don’t think this will happen in my lifetime.
Thank you for reading.