This post is a shout-out to the often overlooked AI models, amidst all the hype around GPT-3 and similar models.
“Oh, now everyone’s fascinated by AI just because it can generate coherent sentences?” they would say, if only these models could generate coherent sentences themselves.
Let’s take a moment to empathize with the underappreciated “Sentiment Analysis” model.
This model has the ability to determine whether text expresses happiness or sadness. It can effectively analyze data from survey questions, customer service emails, or helpdesk tickets. I’m absolutely certain that this model feels neglected and undervalued.
Now, let’s shift our focus to the unsung heroes of “Invoice Processing” and “Receipt Processing”.
Despite being trained on thousands of document layouts and offering cost-effective processing per page, companies still rely on humans or outdated, expensive accounts payable systems instead of building integrated and customized solutions on the Power Platform.
Lastly, spare a thought for the newcomer in town – the “General Document” model. It’s incredibly intelligent!
This model can extract key/value pairs, tables, and checkboxes from ANY type of document without the need for pre-training. It can even recognize handwriting. Imagine the efficiency it could bring to form and document workflows, all at a price of just $15 for 1,000 pages.
Azure Cognitive Services and Form Recognizer are suites of machine learning models built on decades of research. They’re proven, cost-effective, and readily available in Australia without any application process. They’re reliable, safe, and ready for production use.
It’s unfair that these models haven’t received the same level of attention and hype. They must feel a pang of envy towards all the recognition that chatGPT receives.
However, because they can’t write a limerick in Eminem’s style or generate website code from a photo, few people seem to care. Even fewer are willing to utilize them in their businesses.
So, I implore you: let’s give these models a chance, and who knows, your company might greatly benefit as a result.
Together, we can make a difference.