top of page

Getting Personal: Why AI Must Be Uniquely Trained for Maximum Impact



Impactful AI technology must be personalized and specialized, Axios Future reported last month, citing AI scientist and entrepreneur Andrew Ng.

Ng, who pioneered AI technology at Google and Baidu, recently launched a visual inspection platform for manufacturers called LandingLens, knows a thing or two about AI technology. He predicts we’re merely on the cusp of the impact we can realize from AI:


"AI has transformed consumer software,” Ng tells Axios. But if you look at the impact it has had on the broader economy, then candidly, we are just beginning the path of transformation."


While machine learning has overhauled software development, not every industry has realized the same transformative change as quickly.

In fact, a reported 76% of C-Suite executives say they aren’t sure how to leverage AI to achieve their growth objectives – but 84% believe they’ll need to employ AI to achieve their objectives.


LandingLens could provide some insight into the challenges many industries are grappling with. The visual inspection platform for manufacturers enables businesses to curb their tedious defect checking process with human workers and instead allows those humans to simply train the model. Using images of successful and defective products, manufacturers can then label defects and allow the model to develop experiments and, eventually, hone its capabilities to identify defects.


But here’s the key: every factory and every product will need a personalized, trained AI model to realize these results. A custom-trained AI requires not only investment but people with the skillsets to train the AI.


Making AI Accessible

The accessibility of AI isn’t only a challenge at the C-Levels – it requires a new way of thinking and learning from human trainers.

At MIT, a group of students banded together to launch the Machine Intelligence Community (MIC), aiming to “demystify machine learning and artificial intelligence (AI) generally,” according to MIT News. Through Slack, student-led reading groups, and student-led workshops, the 500-student groups strive to make AI tools more easily accessible – to anyone.

According to former MIC president Moin Nadeem, the hands-on experience MIC provides is essential. “Students learn fundamental concepts in class but don’t know how to implement them,” Nadeem told MIT News. “I’m trying to build what freshman would have liked to have had: a community of people excited to do interesting stuff with machine learning.”


Making AI accessible to everyone still appears to be a long term plan. According to a recent Deloitte report, 68 percent of executives reported a moderate-to-extreme skills gap – globally.


One key challenge in closing this gap is how to equip students and professionals with the right knowledge, given academic and training programs struggle to keep up with the rapid advancements in the field, Forbes says.


The solution: expanding the technology and math education provided to give students the foundation they need for on-the-job learning.

At Education Unbound, we tirelessly pursue the inclusion of STEAM (Science, Technology, Engineering, Art, and Math) in education for reasons just like this. We’re on the cusp of realizing transformative efficiencies with AI across industries, but getting there will take a significant number of professionals equipped to train and build impactful AI solutions. Equipping students with this kind of expertise can’t be accomplished with traditional academic teaching – it requires a hands-on approach to learning that will ignite curiosity and the passionate pursuit of learning necessary to stay current on ever-evolving technology.


Visit https://www.educationunbound.org/ to learn how you can join us and help equip underprivileged students to face the jobs of tomorrow with the skillsets and education to make an impact.


Sources:

1. https://www.axios.com/newsletters/axios-future

2. https://techcrunch.com/2020/10/21/landing-ai-launches-new-visual-inspection-platform-for-manufacturers/?utm_source=newsletter&utm_medium=email&utm_campaign=newsletter_axiosfutureofwork&stream=future

3. https://www.accenture.com/us-en/insights/artificial-intelligence/ai-investments?utm_source=newsletter&utm_medium=email&utm_campaign=newsletter_axiosfutureofwork&stream=future

4. https://news.mit.edu/2019/want-to-learn-how-train-ai-model-ask-friend-0625

5. https://www2.deloitte.com/us/en/insights/focus/cognitive-technologies/ai-adoption-in-the-workforce.html

6. https://www.forbes.com/sites/bernardmarr/2018/06/25/the-ai-skills-crisis-and-how-to-close-the-gap/?sh=591c5fd831f3