November 2018 is the last time I blogged about the big, bad “Bots” that many fear will take away humans’ jobs. Given the economic impacts of the COVID-19 pandemic, more businesses are looking for ways to reduce costs, conserve cash and make the most out of investments. Technology investments in artificial intelligence (AI) and machine learning (ML) are predicted to increase.
That can be scary news, but after reading Janelle Shane’s book, You Look Like A Thing, And I Love You, I can offer some additional human-affirming information.
Here’s the big take-away. AI can’t do much without humans!
Shane refers to AI as ‘machine learning algorithms, computer programs that are given data and asked to perform a task with that data.’ However, humans don’t tell the program how to accomplish that task — the AI has to figure it out independently using trial and error. That trial and error approach can discover novel solutions to problems that humans would never consider.
The big downside is that the AI may be working on the wrong problem.
For example, one AI tasked with playing Tetris was told to do whatever was needed, not lose the game. Instead of learning how to play better, when the AI was about to lose, it simply paused the game…permanently. That way, it would never lose!
Bottom line – Humans will always be needed to make sure that AI solves the right problems.
If you are in a role that you think is a prime target for an AI replacement, let me share the types of jobs that humans must fill to support the use of AI:
- Building the AI device in the first place. This job involves anticipating the kinds of mistakes that machine learning tends to make, looking for those mistakes, and even avoiding them.
- Identifying and programming the right dataset from which the AI will work.
- Cleaning up a messy dataset to remove distracting or confusing data.
- Maintaining the data and context as the world and relevant context changes. There is a constant need to adjust algorithms and fix newly discovered problems.
- Detecting and correcting bias to combat the tendency of AI decision making to perpetuate bias.
- Recognizing offensive content and making updates to the algorithm. Ever had an embarrassing auto-correct mishap?
So, what do you think? Do you have the capabilities to tackle this type of role? Is this the type of work that sounds interesting to you?
If this calls for a career shift, may I suggest some retooling and upskilling efforts and become a qualified candidate.
Here are a few options to research:
- MIT Sloan has an online certification program, ARTIFICIAL INTELLIGENCE: IMPLICATIONS FOR BUSINESS STRATEGY. You can gain the knowledge and confidence to support the integration of AI into your organization.
- Coursera has a course, AI FOR EVERYONE. This course is mostly non-technical; engineers can also take this course to learn AI’s business aspects.
- Build talent internally through training and experiences on the job. Invest in learning, use internal and external experts, partner with local universities, launch projects to apply learning, identify the right people for training, and make AI skills development a continuous process.
- Check out these resources for Machine Learning:
- Google offers a free 15-hr machine learning crash course as part of the AI resource center (TechRepublic)
- IT leader’s guide to deep learning (Tech Pro Research)
- Machine learning: The smart person’s guide (TechRepublic)
Remember, there is no ‘man vs. machine’…it’s a collaboration, so let’s get crackin’!
Martha Duesterhoft is a Partner with PeopleResults. Follow her on Twitter @mduesterhoft or connect via email at mduesterhoft@people-results.com.
