AI progress
Blog

A Peek at AI’s Progress

It seems like every organization is working on Artificial Intelligence (AI) right now. Given this sheer volume of activity, it’s understandably difficult for most technology professionals to keep track of how much AI progress they’re making – and in many cases how much they’ve already made.

Fortunately for you (and anyone else paying close attention to AI progress), the latest edition of Stanford University’s AI Index Report has just been released. And this year’s report tracks three times as many data sets as 2018’s version did. In fact, it compares 28 countries’ AI progress activities across more than 30 key performance indicators to give readers the industry’s most comprehensive understanding.

To save you some time sifting through the paper’s 290+ pages, we pulled five important highlights from the report:

1. Research’s Popularity is Skyrocketing

Looking back, AI’s roots can be traced back to the 1990s when computer programs competed against human chess experts. Considering that starting point, it’s hard to believe AI has grown to become today’s most widely researched and studied technology.

However, in the last 20 years there’s been a 300% increase in the number of peer-reviewed papers published on AI. If that wasn’t enough proof that AI research is more popular than ever before, there’s conference and event attendance stats to back up this assertion, too. NeurIPS, for example, is the biggest AI-focused conference in the world – and it expects to have more than 13,500 attendees this year (an increase of 800% since 2012).

2. People Want to Learn More

While AI progress has obviously captured the attention of academia and researchers, global fascination with this technology doesn’t stop there. Enrollment in AI-focused degrees, certifications, and professional training courses is at an all-time high.

In fact, AI is the most popular specialization for computer science graduates across North America – over 21% of PhDs choose this focus today. That’s more than double the rate of the field’s second-most popular discipline: security/information assurance.

3. Despite Recent Developments, the U.S. Still Leads in AI Progress

Though China publishes more AI research than any other nation today, papers produced in the U.S. continue to produce a greater impact on the overall industry. As proof, U.S. authors are cited 40% more than the global average.

The U.S. also invests almost twice the amount of private money into this technology – almost $12 billion compared to China’s $6.8 billion – every year. It also files three times three times the number of patents than second-place Japan annually.

4. Algorithms Are Faster & Cheaper Than Ever to Train

Research is great, but AI progress ultimately counts for little if the technology can’t be developed and scaled. Fortunately for everyone involved, this seems to be the trend taking place.

In the case of training a machine vision algorithm to manage ImageNet – a popular dataset – the time involved fell from approximately three hours in 2017 to just 88 seconds earlier this summer. Costs also decreased throughout the process from thousands of dollars per initiative to just tens of dollars today.

5. Self-Driving Cars Are the Future

Autonomous vehicles are the recipient of more private investment than any other AI-driven field. In fact, almost 10% of all global private growth — $7.7 billion – was focused on this industry.

That said, several other notable areas were included in the AI Index Report’s findings: medical research, facial recognition, robot process automation, and supply chain management also received a minimum of $500 million each.

AI Progress Won’t Replace People Anytime Soon

As impressive as these trends and stats seem, it’s important to realize that this technology is still a long way off from being able to surpass human intelligence. Today’s innovations excel at tasks that feature clear rules and easily simulated training scenarios – and few current workplace tasks can be described as such.

Then there’s the issue of knowledge transfer. With few exceptions, AI systems trained to perform one task typically don’t have the ability to easily apply what they’ve learned to something else yet. While they make exceptional single-use applications, AI is nowhere near ready to be flexible enough to replace stand-in human professionals in most industries.

However, expect improvements and investments to continue to accelerate anywhere AI progress is concerned. The future of technology is an exciting one – and a disruptive one if you’re prepared to take advantage of the efficiencies and performance improvements it has the potential to create.

If you’re looking to implement AI and integrate these innovations into your technology management environment, you need to be at AOTMP® 2020. Fortunately, there’s never been a better – or cheaper – time to register than now!

AI progressClick the image above to sign up for the industry’s premier telecom, mobility, and IT management professionals conference now. If you register before December 31st, you’ll save an extra $500 off your three-day event pass! So, what are you waiting for?