AI: The Ultimate Problem Solver? Costing Billions, Impacting Millions.

AI: Solving Problems, Costing Billions, Impacting Millions.

Big Data is really a Big Deal since we're being already inundated with advice, and it is going to receive a whole lot more overwhelming really, very shortly.

The research company IDC estimates that over two years we'll have now been shrouded in an astounding 40 zettabytes of information (a zettabyte will be 1 sextillion bytes), or even 50 times the information that existed only in 2010. This is going to be the equal of 5,200 gigabytes of information for every single man, woman, and child on Earth.

Nowadays, various bureaus of the national government handle enormous stores of information so as to serve countless millions of Americans.

How can AI improve those agencies capacity to perform their job more efficiently? Let us look at just two examples: the U.S. Department of Veterans Affairs and also the U.S. Patent and Trademark Office (USPTO).

The Department of Veterans Affairs has 21 million active documents, representing every veteran who's served our state.

Somewhere in these documents lie clues which could help people recognize the 22 veterans daily -- nearly one every hour who will commit suicide. But we do not possess the human resources to sift through these documents for the vital indicators related to a greater probability of suicidal actions.

But with weeks-extended waits to see a physician, VA staff could hardly keep up with this vet coming for easy check-ups.

However, AI expert systems are made by trained experts educated about suicidal causes can take action. They can identify veterans at risk, and design interventions which may save hundreds, and possibly thousands, of lives.

Check out the Patent Office. The USPTO has some 14 million records representing the busy shop of patents issued from the authorities for publication, non-obvious, along with inventions that are useful.

These Patent Office documents are called the best storehouse of scientific knowledge on Earth. And that is just what it really is.

Each year, the Patent Office receives more than 600,000 applications for patents, each of which includes a mean of 15 maintains and 75,000 figures.

Every one of those applications have to be individually assessed for novelty, non-obviousness, and usefulness -- that among other things, means assessing them from the 14 million patent documents from the USPTO database.

Though the Patent Office uses 8,500 highly-skilled patent examiners, each of whom has considerable experience in several technical areas, it is a colossal or even borderline-impossible undertaking.

And it is one that every examiner just has a couple of hours and really limited funds with which to perform it. And due to these constraints, many specialists say there has been a decrease in patent quality.

How can AI assist the Patent Office to display out these programs undeserving of a patent, and then make sure just that really worthy creations get a patent? For starters, an AI system may pre-screen and pre-search new programs searching for immediate similarities in their promises to people of already-existing patents, that might consequently spot the ones that appear to be not really first.

Examiners would naturally follow up by using their particular human judgment into the test, however, AI would save time and enhance the examination procedure.

An individual could even envision an AI system which seemed for shared markers of highly-valuable patents -- maybe citation frequency, a portion of non-patent prior art mentioned, or other signs.

In any case, an automatic artificial intelligence program may be a blessing to this patent examination procedure and help raise the quality of patents.

In point of fact, the Patent Office did host a"bake-off" of only these kinds of AI pre-screening and hunting capabilities per year ago.

Some 22 companies engaged, along with the competition was won by groups from the giant consulting company Deloitte, also by a really promising young AI applications start-up named Artificial Intelligence Patents (AIP), correlated with Duke University at Durham, North Carolina.

The AI bake-off has been termed a success. But in the transition to another manager and government in the Patent Office this season, progress has stalled.

The attempt ought to be restored.

Most of Us have a stake in enhancing the job of their Department of Veterans Affairs and also the U.S. Patent Office.

The VA appears after the women and men who risked everything to care for our country's security, and we all owe them a whole lot more assistance than they're getting today.

In terms of the Patent Office, it is the incentive motor of America's ongoing technological greatness along with also the key to our future wealth and our state's global competitiveness.

Yes, telling Alexa to turn our lights play us a tune is fantastic. However, the assurance of AI is far higher -- and a lot more important -- than that.

Let's not squander the chance.