Ideas for Your AI Strategy in 2023

Most likely you’ve heard about the advancements in Artificial Intelligence recently. Like many, you may be unsure how it may be applied to your business. Or you could be looking to expand your AI strategy more in the coming year. In either case, here are some ideas that you can use to leverage AI and help your business in 2023.

Writing

Writing is one of the most obvious new uses for AI. 2022 saw a huge decrease in the barrier of entry for Generative AI (AI that makes stuff). Especially when it comes to so called Large Language Models (LLMs).

You probably heard of ChatGTP and GTP-3 which two of the best known models. However, in the coming year, there will be many more from companies such as Google.

ChatGTP in particular created a lot of buzz because you can have a human-like conversation with it and it is very easy to use.

Some ideas of the things you can write with LLMs are:

• Drafts of blog posts.
• Product descriptions.
• Promotional emails and newsletters.
• Scripts for videos and podcasts.
• Sales emails.
• Customer support answers.
• Books and short stories.
• Computer code.

While you can teach them about your own particular business, these models are mostly “trained” from data on the internet and as such they aren’t always correct in what they create. So right now (February 2022) we don’t recommend that you use the content produced by these LLMs directly. They are best used as a way to help your content management team work faster by producing drafts which are later edited by humans.

Text to Image

Another interesting application of AI is image generation. By feeding descriptions into an AI system, it can generate images that look realistic and can be used for a variety of purposes. The most well known products in this category include Dall-E (by the makers of ChatGTP), Stable Diffusion, and Midjourney.

For example, some of the images AI can generate include:

• Blog cover images (like the one in this article which was generated by Dall-E).
• Logos and icons.
• Product packaging designs.
• Mockups of products.
• Background images.
• Visuals for social media posts.
• 3D models for animation and gaming.
• Educational illustrations.
• Book illustrations.

It can also be used to edit existing images by doing things such as removing objects from a photo, changing the style of the image (for example, making a photo into a cartoon), changing parts of an image to something different, or filling in details that don’t exist.

We like to use image generation AI to make interesting illustrations for content that we wouldn’t otherwise have a designer create, such as the image for this blog article.

The ethics of generative AI is still hotly debated. Especially for image generation since the models are trained from images “copied” from the Internet (usually without the artist’s permission). So we don’t recommend directly selling the output of these AI without editing just yet.

Enrichment

Enrichment is another interesting use of AI. For example, AI can be used to automatically add tags, metadata, and other information to content. This can be done by analyzing the content and making decisions based on what it finds.

It can be used to automatically add tags to articles, images, videos, and other content. It can also be used to automatically create summaries of articles, find keywords and topics, and even to create structured data from unstructured content.

For example:

• Categorize objects and faces in photos.
• Sentiment analysis (we use GTP-3 to automatically detect the mood of customer service emails and chat messages).
• Summarizing articles or reports.
• Making images and audio searchable.

AI can also be used to determine the quality of content. For example, AI can be used to measure readability, grammar, and other metrics that can help you determine which pieces of content are the most effective.

Monitoring

AI can also be used to catch issues faster. Some of the most common uses of AI in for this purpose in 2023 include:

• Anomaly detection to check if a measurement is different than usual. For example, sales, website traffic, or errors
• Line monitoring using machine vision to see in real-time if the parts being built defect free.
• Quality assurance to check output.
• Monitoring the sentiment/mood of the discussion around your product.

Customer Service and Sales

While just about no one likes talking to a robot when they are having a customer service issue, AI can also be used effectively to help customers faster and also decrease the burden on customer service representatives.

Besides writing new content, AI can also be used to:

• Answer customer service questions automatically before escalating to a human.
• Searching knowledge bases and providing targeted answers.
• Recommending products or services based on customer behavior.

Some of these uses of AI have been around for years. But now, with advancements in technology they are available to a much wider audience.

Looking to 2023 and beyond

When thinking about your AI strategy for 2023 it’s best to think about it as an assistant to help your team be more effective. Not as a replacement for team members. AI can help create content, enrich data, monitor your systems, and take care of routine customer service tasks. But your team are the experts, you know best.

In 2024 and beyond AI may be able to take the place of more tasks completely. But until then, you shouldn’t discount what it can do to help you and your team to be more efficient.