A Brief History of Artificial Intelligence
What is AI?
Artificial Intelligence (AI) has been getting a lot of media coverage lately, and for good reason. AI applications are quickly changing the way people interact with each other and with the world. Many people use AI every day without even realizing it. If you’ve used a GPS app to find the best driving route to work, accepted a product suggestion that an online store made based on your past purchases, or asked Google Home or Alexa to play a song for you, you’re a consumer of AI.
Sounds pretty simple right? There’s also quite a bit of misunderstanding about AI and what AI applications can and can’t do. This is usually based on perceptions we have from science fiction books and movies. (Are computers suddenly going to become conscious and try to dominate the world? Not for the foreseeable future.) So let’s start by taking a look at how an AI application really works:
A screenshot of what AI looks like in Salesforce. At right, see how Einstein has predicted a college student’s likelihood to stay in school (for example).
In general, AI applications (like Salesforce Einstein) do a few things:
- Discover insights by comparing some type of input with data they already have
- Do statistical evaluations to predict outcomes
- Recommend next steps based on the results of their evaluations
- And some AI applications will automatically implement actions themselves via automated workflows (such as sending an email every year when it’s someone’s birthday)
The term “artificial Intelligence” is used to describe these applications because their processes mimic the way our own brains make decisions: If you get cold from standing outside in the snow, then the next time you look out your window and see snow, you’ll know that it’s probably cold outside and put on a coat instead of a pair of shorts before you go out. In much the same way, an AI application interprets text, sounds, images, location coordinates or other types of data, compares it to data that it already has and then decides what to do next. The only difference is, the AI application might need to have 10,000 or a few million examples rather than one winter morning to be “trained!”
Two common AI examples you may have already interacted with are:
- Chatbot applications that appear to have human-like conversation skills. These evaluate text inputs and compare them to a set of known values, along with potential misspellings, alternate phrases, etc, to identify keywords and possibly even the mood of the speaker.
- Predictive Analytics Applications use statistical modeling to find cause-and-effect relationships in large sets of data and use the results to try to predict future events.
Once an AI application completes its evaluation, the next step is to determine the best way to respond. This can be anything from answering a question, forecasting future trends for an organization, stopping a self-driving car at a stop sign, or any other action that the application was designed to carry out.
Artificial Intelligence isn’t new. I took my first AI class in graduate school all the way back in 2003, and even then, most of the concepts I just described had already been around for decades. The reason AI applications are becoming more prevalent today is because the amount of computing power that they need to function is finally starting to become readily available. This looks like:
- Millions of calculations to accurately identify patterns
- Massive amounts of storage space to have a big enough set of data to ensure that predictions are accurate
- Determining how to respond to an input usually requires an application to build decision trees that evaluate billions of possible responses and then determine what the potential outcomes might be for each of them before choosing the best one
- Internet bandwidth is getting faster, from copper wires to fiber optic cables, enabling more computation to be done on powerful, faraway servers where you can access the results of intensive calculations easily on your lightweight laptop
To make an application truly seem “intelligent,” all of that number crunching has to happen within fractions of a second. Most older computer hardware simply wasn’t up to the task. While scientists waited for it to catch up, they continued to evolve their ideas about how AI could be used, and eventually their theories started to become reality.
What AI means for Nonprofit and Higher Ed
So, what does this all mean for nonprofits and education organizations? Quite a bit, actually. Some organizations are already using AI to improve fundraising by allowing their applications to automatically score leads and craft appropriate, personalized emails to potential donors. Others are using or developing AI applications to detect diseases or to place refugees in locations where they’re most likely to find work. In the higher ed space, AI applications are being used to identify coaching methods to prepare at-risk high schoolers for college, predictive analytics are being used to identify students who are at risk of withdrawing from their programs, and chatbots are being used as communications tools to help improve the student experience from recruitment through graduation.
AI investments are going to continue to increase in coming years, which will be particularly beneficial for nonprofits and higher ed organizations that have a limited number of available resources. AI can help these organizations connect with their constituents in ways that eliminate a lot of manual effort while still providing a personalized touch.
When you want to innovate, it helps to have a diverse team. Photo of Amplify members at a Salesforce.org event.
So what do you need to do to bring AI into your own organization? Follow these steps to get started:
Step 1: Learn more about AI on Trailhead, such as Artificial Intelligence Basics, Salesforce Einstein Basics, or dive in to Deep Learning and Natural Language Processing. More links are below for specific resources for nonprofits and higher education institutions.
Step 2: Develop a vision for what you’d like to accomplish. Imagine having a personal assistant whose job is to understand your organization’s goals and suggest ways to meet them. Now imagine that your assistant also knows what to do next, continues to learn from their actions and can make adjustments to their plans based on their learnings as they go along. What would you ask that person to do? And what information would they need to know to be successful? Write down your thoughts and use them as a starting point.
Step 3: Once you have a vision in mind, reach out to Salesforce.org or one of our partners to discuss what’s needed to bring your vision to life.
Resources to bookmark for information about how Artificial Intelligence can help your organization
- The AI Week Nonprofit Digital Magazine
- E-book: AI for Nonprofits
- The AI Week Higher Education Digital Magazine
- E-book: Education Cloud Einstein
- AI Week overview for Higher Ed
- AI Week overview for Nonprofits
Nonprofit AI Webinars and Master Classes
- Nonprofit AI Broadcast: Social Good in an AI-Powered World
- Webinar: Intro to AI for Nonprofits
- Master Class: Drive your Mission Forward with AI – Best Practices for Smarter Impact on Wednesday, March 20 at 9 am PT / 12 pm ET
Higher Ed AI Webinars and Master Classes
- Broadcast Event: Higher Education in an AI-Powered World with Cal State East Bay
- Webinar: Artificial Intelligence: How Can it Help Higher Ed?
- Webinar: AI for Advancement
- Recruit Smarter with AI: 4 Ways to Unlock Admissions Data to Meet Enrollment Goals on Wednesday, March 20 at 11 am PT / 2 pm ET
- Master Class: AI for Every Advisor: 3 Ways to Maximize Student Success with Education Cloud Einstein on Thursday, March 21 at 11 am PT / 2 pm ET
You Might Also Like
An Interview with Lindsay Gruber on building corporate resilience through pro bono volunteerism.
Each year, as we observe the International Day for the Abolition of Slavery on December 2, we have an opportunity…
A conversation with Jennifer Lotito, the President and COO of (RED), to learn more about (RED)’s success model and how…