3 mistakes people make when thinking about AI

Adrian Goergen

Introduction

For most of my life, I fell victim to some conventional wisdom about artificial intelligence. I soon realized that many other people fall into the same traps. I’ll uncover the three most common mistakes people make when thinking about AI here so you don’t have to waste the same amount of time.

Mistake #1: AI is about futuristic robots

Artificial intelligence is not all about building super-intelligent algorithms and human-like machines that outperform us in any discipline, even if science fiction movies have popularized this image time and time again.

Ever seen the movie Her where a lonely and depressed man fell in love with an intelligent virtual assistant? Ex Machina tells a similar story where a humanoid robot manages to escape from its test space by manipulating the feelings of a programmer. Captivating 🍿 but far from reality.

AI is less glamorous than what Hollywood movies and media headlines suggest. But it is still pretty cool what it can do and what it already does for you every day:

  • The spam filter in your email inbox
  • The movie suggestion algorithm of your Netflix account
  • Your phone's autocorrect feature

Artificial intelligence is an area in computer science. Simply put, it refers to machines performing tasks that normally would require human intelligence. There are three subfields of AI:

  • Weak AI (also known as narrow AI): less intelligent than humans, algorithms limited to a specific problem
  • Strong AI (a.k.a. general intelligence): as intelligent as humans, algorithms can apply intelligence to any problem (“general”)
  • Super-intelligence: smarter than humans in all domains

Hence, artificial intelligence is not only about super-intelligent robots – those would belong to the last category and are a far shot from where we stand today. Most models of artificial intelligence currently fall under the first category, weak AI. These can be at the core of some incredible applications, too – but they don't make a good Hollywood blockbuster.

Mistake #2: AI is expensive

While artificial intelligence is on the agenda of many business executives, an often-cited excuse is that it’s too expensive to introduce. The term artificial intelligence arouses the picture of a highly complex and expensive to develop a computer system – out of reach for most CEOs.

As described earlier, this might be the consequence of a distorted picture of AI through science fiction movies. Movies like Her and Ex Machina suggest artificial intelligence to be something too complex to implement in an average company.

Also, business executives might underestimate the recent developments in AI, which saw costs decrease tremendously. AI seems to still be a privilege of rich tech companies.

While true in the past, this view is outdated nowadays. Here are a few examples of how companies of any size can implement artificial intelligence in their business at little costs:

  • Integrate a customer service chatbot (using e.g. LivePerson, GetJenny)
  • Automatically assess the suitability of job applicants using AI-based screening software (e.g. ideal, pinch, CVViZ)
  • Improve personalization and timing of email marketing campaigns (e.g. Seventh Sense, Automizy)

Summing up, AI does not need to be expensive. But ignoring AI might very well be, at least in the long-run – and there are good reasons why that is the case.

Mistake #3: I need to know how to code to utilize AI

I agree: Developing AI models sounds like a task for nerds, those people who have started to code when they were children. However, the truth is that you don’t need to know how to code to develop an AI model. What sounds utopic has become reality thanks to recent technological developments.

Do you remember the times when developing a simple website required some serious coding skills? Today, any person can build their own website without a single line of code. Companies like Wix, Webflow, and WordPress provide you with design templates and guidelines. All you need to do is shift around boxes, insert pictures, and write text.

What works for websites begins to work for AI models as well. Similar to Wix or Webflow, there are firms out there that allow you to easily build AI models without coding.

To use one of our own examples, imagine you want to build an image classifier. At colabel, it is enough for you to come up with an idea and data. Similar to building a website with Wix or WordPress, it is more about understanding the workflow than actually writing the code.

Instead of writing code, what you actually need to do is the following:

  • Define your problem: Make sure you exactly know what task to solve beforehand. A simple example would be to label dog and cat pictures accordingly.
  • Get & label data: You need a set of “correct examples” to train your model, i.e. dog pictures associated with the class dog, cat pictures with the class cat. A perfect task for your intern 😉
  • Train the model: Feed the model with more pictures of cats and dogs.
  • Improve the model: Sometimes your image classifier is not 100% confident. You’ll have to manually double-check the corresponding classification. Doing so, your model will improve accuracy over time.

This is all you need to do! Now, your image classifier is ready to the work for you. Did you notice that not a single line of code was needed? In case you now want to dive deeper into it, we wrote a more detailed guideline on how to build an image classifier in no time and without code. Or just sign up and build your own!

Conclusion

To summarize the common misconceptions about AI:

  1. AI is not only about futuristic human-like robots. Most of the current AI applications are within the field of narrow AI. They are powerful yet less glamorous.
  2. Integrating AI into a business need not be expensive. Not integrating AI certainly is (in the long-term).
  3. You don’t need to know how to code to utilize AI. You can start building your custom AI model today.

I hope that we could clear up some misconceptions about artificial intelligence. In case you have any questions about the article or any wishes for new content feel free to reach out to us via Twitter or LinkedIn. We are happy to help!

Share on social media: 

Now that you're here

colabel is a tool that allows you to train AI models on images, documents, and text data. You can rebuild manual workflows and connect everything to your existing systems without writing a single line of code.

If you liked this blog post, you'll probably love colabel.

Thank you! We'll be in touch soon.
We are sorry - something went wrong. Please try again!

More from our Blog

Human-in-the-loop in machine learning: What is it and how does it work?

Human in the loop (HITL) is a well-known and powerful concept for reaching outstanding collaboration and performance in artificial intelligence.

Read story

5 easy ways to save money using AI in your business

Almost any company can leverage the benefits of AI – here are five ways that can be implemented rather easily.

Read story

How to leverage AI in your business - without coding knowledge

SMEs can leverage AI technology for productivity gains without coding knowledge - and tap into productivity gains & cost savings. Here's how.

Read story

Stay inspired.

Sign up and get thoughtfully curated content delivered to your inbox.
We will never share your email address with third parties. And we don't like spam either.