The hype around AI is in overdrive with hugely inflated user expectations. Unfortunately, such sky high expectations rarely meet reality.
DigitalBiz has created two new videos to help owners bypass the hype and understand how to combine Applied AI (pre-trained on industry data and usable “out of the box”), Generative AI (like ChatGPT which can create original content ) and human experience – to maximize productivity gains.
For additional insight and to view the videos, visit https://digitalbiz.ai/
The costly pitfalls of using the wrong artificial intelligence solution in business can take the unwary adopter by surprise – if they are taken in by the hype around machine learning, deep learning, natural language processing, neural networks and a host of other issues.
For example, the incredible capabilities of ChatGPT and other generative AI models have been described as magical by users. But if used incorrectly it creates more problems than it resolves.
When used correctly, AI is delivering average productivity gains of 40% plus. Companies ignoring AI will get left behind and have little choice but to get involved. So to get to the productivity stage, it is important to understand what AI technology is and what it is not.
Artificial intelligence is just computers following human rules – which have to be coded. As the AI revolution has only just begun, most rules won’t be coded for decades, because data science is playing catch up. For example, of the 25,000 genes in the human genome only 2% make up people’s DNA. There are no rules for 98% of genome genes. And in the absence of rules, computers guess!
Take ChatGPT. Text produced is just data in the form of words strung together using rules. If ChatGPT is asked “Describe what genes in the human genome do” the result is impressive sounding conjecture combined with errors mixed with truth. ChatGPT is not smart enough to know if source material is accurate, relevant or a waste of time. It just throws out word combinations.
Here’s what happens when things go wrong.
In 2017, the LA Times reported a 6.8 earthquake in Santa Barbara. A large quake always makes the news. So did this one. It made the news in 1925 when it happened! The AI spotted a US Geological Survey 1925 data revision, and logged it as a 2017 event. The LA Times published without checking!
The lesson for all users is AI is not yet ready for prime time. It is still deep learning the rules and human oversight is needed for most applications. Applied AI is brilliant at executing data driven, repetitive or predictive tasks quickly. Generative AI like ChatGPT creates original content. Both need human oversight.
For the next few decades at least, the future is combining the best of AI capabilities with the best of human experience.
The following worked example combines Applied AI, ChatGPT and humans to tackle common content planning and operations pain points.
Problem: Ongoing demand for content is increasing but there’s no plan. Hot topic questions searchers want answers to need to be identified; hot topic gaps in existing content need plugging; new content needs to be created and distributed.
These tasks are data driven, repetitive and predictive. Applied AI automates them all. It content plans, identifies hot topics and content gaps, and distributes.
The missing link is creating factually accurate, relevant hot topic content with emotion and understanding which genuinely connects with people. This is where ChatGPT comes in.
Getting the right content from ChatGPT requires clear and concise instructions. So ChatGPT experts have created prompts – pre prepared templates – which anyone can adapt to trigger a desired response.
For example, “Write a 700 word blog on networking for store owners” generates a different response to “Write a blog on networking”
ChatGPT creates content but lacks emotional understanding until humans edit it. It then learns from the edits and with every revision gets closer to developing content that genuinely connects with people.
Use the right prompt, get the right answer, and adjust for human consumption. ChatGPT – with human help – develops the appearance of emotion and understanding. It is not the “emotional intelligence” Daniel Goleman would recognize, but it’s the beginning of Applied Artificial Emotional Intelligence. And it maximizes reader engagement – which is the goal.
Here is the final process which combines, Applied AI, Generative AI and Human experience:
The applied AI creates a plan of hot topics to write about. Hot Topics feed a ChatGPT prompt. Unique content is generated for editing by humans. With each edit, ChatGPT soaks up the writer’s experience. When enough edits accrue, ChatGPT is able to mimic the writer’s tone, style and emotional understanding for this article and others in future. The Applied AI processes the content, publishes to multiple platforms, and consumers engage.
Adrian McKeon, DigitalBiz CEO states ” Don’t rush into the AI waters. Instead, contemplate what mix of AI technologies could improve everyday processes. If your task requires data input, is predictable or involves prediction-making, it’s highly likely that AI can do it more efficiently and economically. Start with a trial run to see if you encounter any snags; if all goes well, keep building from there!”
With DigitalBiz, small businesses can unlock the power of artificial intelligence and get access to content planning, creation and distribution high-tech tools usually available only for big corporations. Its comprehensive AI content management service is accessible, cost-effective, and effortless to use.
Interested parties can access the newly launched videos at https://digitalbiz.ai
Initial Business Centre
207 Regent Street
Disclaimer: The views, suggestions, and opinions expressed here are the sole responsibility of the experts. No Sahyadri Times journalist was involved in the writing and production of this article.