5 things AI does better than humans

Artificial Intelligence (AI) does better than humans in many areas. It is getting more advanced each day. Examples range from Boston Dynamics demos backflipping robots to the likes of AI is dominating advertisingBut it doesn’t stop there so let’s get a better sense of what AI can do better than humans in 2019.

1. AI Does Better In Gaming

When AI beat the reigning ‘Go’ champion Lee Sedol in 2016, researchers rejoiced. Furthermore, in 2017, a new neural network –AlphaGo Zero’ beat the old ‘Alpha Go’ 100 times in a row. Moreover, it also taught itself how to play using only the basic rules of the game. Even though ‘Go’ is far from the only game humans play, there are new AI’s in town playing all kinds of games.

The computer science department at Carnegie Mellon University developed an AI named ‘Libratus’ that won in no-limit Texas Hold ‘Em game. What is impressive here is not so much that it won at Poker, but the type of Poker it won. In 120,000 hands, ‘Libratus’ defeated four Hold ‘Em experts in what is known as an “imperfect information” game.

Another example of an imperfect information game where AI is schooling us humans is OpenAI’s Five’ that beat top professionals in Dota 2.

As time passes the list is growing. It will not be long for the list to be complete where all world champions will be AI.

2. AI (potentially) Does Better AI

Machines have built machines for a long time. Therefore it isn’t unreasonable to think that AI could build better AI than any human could. In theory, AI is able to pull this off, but it isn’t certain if it can be done in reality.

Both Google’s AutoML and Microsoft’s DeepCoder hold an unprecedented capacity to build the next generation of AI. DeepCoder does not simply copy the building blocks of code as researchers have given it to it. The algorithm looks at how the codes fit together, how they function and it learns and recognizes other codes.

The video below, shows Google’s CEO telling his team “we have to go deeper”, describing AutoML. This is also a nod to the movie ‘Inception’. 


3. AI Does Better At Providing More Accurate Medical Diagnosis

Ever since the achievements of AI has been the talk out of town, a major focus area for their use has been in medical diagnosis.

For example in the field of oncology and diagnosis of cancer, it is challenging for humans to have an absolutely accurate diagnosis. The delivery of results by humans had a correctness level of 50 per cent whereas AI reached 90 per cent.

More recently, IBM has used Watson’s ability to absorb huge volumes of information. That will help them with diagnosing rare illnesses. Some of which most doctors may only see within a few cases in their lifetime. It will help doctors at the Centre for Undiagnosed and Rare Diseases at the University Hospital Marburg, Germany, deal with the thousands of patients referred to them yearly. As well as with the thousands of pages of medical records that are supplied by the patients to be analyzed.

Authors including Vinge and Kurzweil have postulated that AI technology will develop up to a point, called “the singularity”. It is when the combined problem-solving capacity of the human race will be overtaken by AI. However, this is not the case for many years to come.

4. AI Does Better At Transcribing

AI can now transcribe audio better than humans. Microsoft’s researchers have been tweaking an AI-based automated speech recognition system so that it performs as well as, or better than, people. They have proven that through testing.

During a test implemented by the National Institute of Standards and Technology, this AI system had an error rate of 5.9%, the same as human transcribers that Microsoft hired. When the test was replicated, it had an error rate of 11.1%. This was almost matching the human score of 11.3%. This shows us that AI leave little room for error in audio transcription and is certainly not any worse than we are. 

5. AI Does Better At Creating Entertainment (than most people)

French songwriter Benoît Carré has collaborated with an AI program called Flow MachinesThis was used to create a Europop album that exceeds expectations. Carré who has worked with some of France’s biggest names like Johnny Halliday and Françoise Hardy explains that the system has the capability to write original melodies. He goes on by saying that the AI can even suggest the chords and sounds to play. However, Carré admits that AI produced music still needs a human touch to bring it all together. So while it is still a very ‘human album’, there is no denying that the future of music will include lots of AI contribution.

Not long ago, MIDI seemed new and futuristic before it transformed music and pop culture for good. It is not a fantasy that AI does better at making music than us humans.

An AI system capable of creating original pieces of art has been developed by a team of scientists. The idea is to make art that is “novel, but not too novel”. 

The researchers from Rutgers University and Facebook’s AI Lab came together to build a system. It sees two neural nets interact with each other to continuously improve their results. One creates a solution or image, the other judges sit in a continuous loop. This is done until the desired outcome is achieved.

The discriminator network was trained on 81,500 paintings to build a knowledge bank. The point is to see what is classed and not classed as works of art. Among those things, AI has also started to write scripts for movies


Conclusion

AI having a real function in everyday life has been displaced from being a thing of lofty dreams to reality. Just as in the past decade we have seen Artificial Intelligent Assistants go from a novelty feature on mobile devices to become ‘Home Assistants’ that run our households and lives. In the next decade, AI will probably become a completely normal tool found in many aspects of our lives. Even though today it might seem far fetched or far off.

AI has already proved it can beat us at games, make art and music. Moreover, it can reproduce itself and consolidate medical records so it can help make diagnoses. It also has the ability to transcribe audio. 

Such as the future of AI is sure to be exciting, so is the future of many other technologies. Therefore, keep an eye on our blog to not miss out on any!
Also, if you have an idea for a native of a hybrid app or a web application, don’t hesitate to contact us! Let’s make the world digital together.

 

Recent Comments