How Will We Use Artificial Intelligence In The Future?
In analysis / By Mark Flynn / 07 August 2018
It has once again been a big month for AI with some really interesting things going on that are vitally important to how we will use AI in future.
The most exciting piece of AI news came out of IBM when they revealed 'Project Debater.'
Now to some, it might actually sound like your worst nightmares coming to life because Project Debater can argue back, however it actually displays a range of impressive innovations.
Firstly, it can understand the context of what someone is saying during a debate, creating its own argument in real time.
AI assistants based on innovations like Project Debator will deliver exciting opportunities for every business.
Now AI assistants aren’t new with many people using the likes of Siri and Cortana, however with Project Debator capabilities have been added to improve how it interacts with people turning it into a really powerful business tool.
In Project Debator’s case, the AI is loaded up with a large number of research papers around a subject and then let loose to come up with arguments for and against the issue.
So it might be time to start thinking about how you'll embrace AI and develop your 'Bots and Bodies' strategy, one which has a core set of people supported by AI assistants making quicker, better business decisions and delivering a first class service.
Perhaps arguing with an AI might not be your idea of a good time but you most certainly wouldn’t want to work alongside Norman the psychopath AI. Norman was designed to be a psychopath from the get-go as he was fed data from the darkest corners of the internet.
Where normally an AI would identify a Rorschach test as flowers or generally happy scenes, Norman could only identify them as murder scenes or images of violence.
This experiment by the Massachusetts Institute of Technology was an attempt to illustrate a growing problem within AI. Bias is something that AI is supposed to eliminate but the truth is that an AI can be just as biased as a human being.
If the AI is fed data that has a bias then the AI will pick it up and any answer it gives will have that bias going forward. An example of this is an AI that was trained on Google news that became sexist and claimed a woman’s place was in the kitchen(!)
These examples show that we can’t expect AI to be automatically right and that we need to monitor the data it is being trained on.
Of course, once you’ve got an AI and you’ve made sure you’ve got good data available, there is the question of how you use it to your advantage.
An interesting example of the many ways you can use AI in business is demonstrated really well by Starbucks. When they launched their reward program and mobile app they ended up with data from millions of customers to help improve their services.
One example is AI’s ability to predict things it thinks the customer might like based on previous buying habits. By comparing what other customers bought, it can come up with new suggestions for future purchases and then alter that list depending on the weather.
That use of AI extends to opening new stores and putting products on shelves.
When Starbucks look to open a new store location, AI tools are used to evaluate the area before going ahead. By looking at traffic patterns and local demographics and their habits in store, they can compare it to other locations and decode if that site would be a success. AI can also help the business to simulate what would happen to other nearby Starbucks stores if that one was to open in that chosen area.
For some, it might be difficult right now to foresee how AI is going to change your business, so let me provide an example.
Is anyone reading this who doesn't own a smartphone? Can you function in your personal and work life without your smartphone? That is exactly how important and intertwined AI will become in our lives.
So start planning right now how you will use AI to change your business, kick-start your thinking with these seven examples of how you might use AI.
- Analyse all previous deals won/lost to help improve forecasting, accurate profit estimations per deal and more accurate close dates on deals
- Support staff retention activities by pulling together data from exit interviews, performance management, surveys and discussion threads to build a real-time picture of the health of your business
- 1-2-1 conversation with your customers is difficult, so use AI to process the massive volume of data you hold on them to help create a hyper personalised service for your many customers
- Analyse video interviews that scan facial expressions to check the suitability of a candidate
- Automate anything that is repetitive or rules based e.g. compliance or staff onboarding
- Automate anything that is time critical or reducing time spent manually matching payments and invoices
- Anything that is error prone e.g. those massive contracts you have to complete that have lots of different parameters you have to set
For more information on anything in this article, please contact me.