If you're going to use AI in your work effectively, you need to know what AI is good at.

If you use AI for something it's not so great at doing, your results won't be so great either, in either quality or efficiency. And if you're not aware of what that AI is good at, you might miss some big opportunities.

We often think about using AI to replace human work or speed up tasks we're already doing, like data entry and searching the Internet for answers. AI can absolutely speed up a lot of human work; however, AI's strengths don't perfectly align with what people are great at doing. It's therefore helpful to think about what tasks are uniquely suited for AI. Some of these tasks we don't do frequently ourselves in our work because they're just too time-intensive, but for AI, it can be a breeze.

In this lesson, we'll discuss the 10 things AI is great at.

The 10 things AI is great at

1. Classification

Classification is probably the benefit of AI you've been receiving for the longest time. If you've used email over the last 20 years, you've noticed that it's gotten really good at classifying emails as spam or not spam. It works so seamlessly in fact that you might not bother to check your spam folder anymore.

Classification assigns data into one of multiple available categories. For example:

• Classify an email spam or not spam

• Classify an expense for travel or marketing

• Classify a lead in the real estate or finance industry

It's a powerful tool and AI is great at it.

2. Extraction

Extraction is pulling out specific elements from some data. AI can usually do this efficiently and reliably. Common data sources are text, files, webpages, datasets, and images.

For example:

• You receive an invoice and you want to extract the due date, amount, and vendor name to add it to your order tracking system

• A photo of a driver's licenses is uploaded to Google Drive as part of an application process, and you want to extract the full name and addresses from the image for your records

• You're scraping pricing pages and you want to extract prices from them for competitive research

3. Summarization

Summarization is taking a larger piece of text and condensing it into a smaller piece of text. Summarization makes our lives easier by making hard-to-consume pages of text into something more digestible. And AI's really good at it!

For example:

• Take a transcript of a 60-minute meeting and summarize it into a 1-pager of notes and action items

• Take a 30-minute podcast and summarize it into show notes

• Take an 8-minute YouTube video and summarize it into just the description and chapters

4. Writing

The fourth skill AI is great at is writing. For example:

• Write a social post based on a content idea

• Draft a proposal based on the transcript of a kickoff meeting with a client

• Write a full video script based on a brief

However, AI writing is a double-edged sword. The results will only be as good as the inputs. If you give AI a bad prompt, poor instructions, and generic content, you're going to get AI slop.

But if you craft a great prompt, a good brief, and give really good examples of the kind of writing you expect, you can get some amazing content quickly.

In addition to including examples in your prompt to show the AI what you're going for, writing tasks also can benefit from optional knowledge source reference materials. Giving the AI access to the last 10 emails you wrote, or posts you made on LinkedIn, or a company blog post style guide, can help improve the quality of a generated email, social media post, or blog post.

5. Media creation

Media creation is one of AI's more recent advances. With media creation, beyond just text, AI can generate images, audio files, and videos. It's a total game-changer for marketing teams in particular. For example you can:

• Create images to add to a marketing email or flyer

• Generate the audio voiceover for a video you're recording

• Generate a full video for an ad campaign

The quality of media creation varies a lot. Some things that you think might be easy—say, creating a simple poster with a few words for an upcoming webinar, or for some reason, drawing fingers—are hard for AI. These tasks likely won't turn out well without careful prompt curation or resorting to post-processing. But other tasks that sound like they should be hard, like creating a large realistic cinematic scene, are easier for AI to generate. This is a great example of AI's strengths not perfectly aligning with people's capabilities.

That said, media creation is getting better very quickly. Explore it now and keep an eye on the capabilities if you have a use for non-text generated content like voice and imagery.

6. Research

AI is amazing at research. It can scour multiple sources to answer a query and bring the results back in one place in seconds without you opening a browser. Our phones have become research assistants in our pockets.

With AI's research capabilities, you can quickly:

• Research and write a dossier about a person you're about to meet for the first time

• Understand your competitors' pricing plans or offerings and how they have changed since you last looked them up

• Deep dive into news about a particular topic

7. Synthesis

This one you might not be as familiar with, or doing on a daily basis. Synthesis is a practice that is extremely valuable to businesses, but we often don't take the time to do it because it takes so long to do manually. With AI, that has changed.

Synthesis is combining multiple strands of information into one synthesized whole. AI can easily take different sources that would be time-consuming for a person to review and combine and summarize the results into a single useful output.

For example:

• Given the customer meetings you had over the last week, synthesize what customers are saying about your product

• Given the support tickets that came in over the past day, synthesize common trends or patterns

• Given the set of announces or launches from a competitor, synthesize their product and marketing strategy

You can imagine how these insights, that you can set up to happen automatically and regularly, often for mere cents or less, could transform how you approach company and product strategy, marketing, customer success, and engineering prioritization.

8. Analysis

If you give an AI a dataset, it can often come up with strikingly good conclusions about what's going on in the data. For example, you can have AI:

• Write your weekly report about sales data and trends

• Analyze why customers have churned in the past week

• Find competitor mentions in recent sales calls and analyze competitor weaknesses and strengths

In all of these cases, it will not only be able to pick out individual insights in the dataset, but also it can create a useful commentary about why a trend is happening in the data or what the implications might be for you. Like you would ask a research analyst on your team, you can instruct the AI what you need to know to make the analysis as useful or actionable as possible.

9. Grading

Like any good teacher, AI can come up with a letter grade or a score for a piece of data based on a set of criteria, as well as some rationale for the score, and if desired, recommendations to improve the score. You can use grading for use cases like:

• Grading applications to a partner program and recommending whether to accept and why

• Grading the likelihood of a customer to renew a subscription based on their product usage

• Grading the quality of your own writing in documents or emails

If you're using grading tasks to evaluate applications or someone's work, and you're taking automatic actions based on the result, chances are that your task will benefit from a human in the loop. Have a teammate review the grade, check for mistakes and bias, and make changes if necessary before any high stakes action is taken. This is why including a rationale along with the grade is so helpful; it gives context to the number and speeds up a human reviewer's approval process.

10. Coaching

Last but not least, AI is a great coach. If you wish you could be a better sales rep, marketer, people manager, fortune teller, you name it, if you have a good dataset, AI can give you recommendations to improve. For example, you can have AI:

• Review transcripts of your last 10 sales calls and give coaching tips on how you can be a better sales rep

• Review the content and reactions to your recent social posts and coach you to write better hooks

• Review the transcripts of your internal meetings over the last month and coach you to be a better facilitator

You may have thought of an AI agent like a virtual intern who can crank out mechanical tasks on your behalf. But with skills like coaching, AI can also play a strategic role in your team's processes or your career goals.

Now you know

Now you know the 10 things AI is good at, and you have 3 examples of each skill:

1. Classification

2. Extraction

3. Summarization

4. Writing

5. Media creation

6. Research

7. Synthesis

8. Analysis

9. Grading

10. Coaching

We should all be finding more ways to use AI in our work. Thinking about how you can apply these 10 skills is a great place to start.