Connect with us

Artificial Intelligence

Elon Musk fears AI: Interesting facts you probably didn’t know about AI

Published

on

One doesn’t need to be a tech expert to have heard about artificial intelligence. The term has become quite common in our everyday life, and this shows how quickly the technology is being adopted. 

The general understanding of AI basically entails computers or machines that are super advanced and smart enough to learn on their own. While this understanding is correct, AI is not as simple. However, for the population that isn’t tech-savvy, it will suffice. 

The complexities of AI might make it sound rather boring to a lot of people, so the idea of AI people have gotten from movies helps keep them interested in the technology. It is safe to say that the first images that come to people’s mind are robots, terminators and more robots. 

To further amplify interests in AI, here are some facts you probably didn’t know about the ubiquitous technology.

Elon Musk is afraid of AI

Tesla CEO, Elon Musk, is no stranger to AI. Tesla cars have semi-autonomous capabilities, so, what is he so afraid of? For someone who plans to colonize Mars, build SpaceX rockets that could get anyone to anywhere in the world in under an hour, no tech should frighten them. 

Well, Musk is genuinely worried about AI. In 2014, at the MIT Aeronautics and Astronautics Department’s Centennial Symposium, Musk unequivocally warned about the dangers of artificial intelligence. His words were, “With artificial intelligence, we are summoning the demon. In all those stories where there’s the guy with the pentagram and the holy water, it’s like, yeah, he’s sure he can control the demon. Didn’t work out.”

In his analogy, AI is the demon humans think they can control, but will most definitely fail if there are no restrictions. Subsequent facts will point at the possibility of an uncontrollable AI. Musk emphasizes the need for more regulations in the creation of AI. Humans are exploring the possibilities when it comes to AI, but a lot of caution needs to be taken. Coming from someone who, needless to say, has no phobia whatsoever when it comes to tech, we’d say we’ll do well to listen to him.

Moravec’s paradox

There’s a paradoxical phenomenon about AI. For a technology that has thrived on logic for the most of its existence, an illogical truth about it makes it all the more interesting. 

The Moravec’s paradox is an observation by AI researchers that making computers perform high skilled jobs such as reasoning is easy to compute but making them perform skills considered easy or simple are harder to compute. Sensorimotor skills such as walking are considered simple, they come naturally to humans. The truth about these skills is that it takes humans several years to perfect. 

They are considered simple skills because we do not actively learn them. It takes a while to learn to walk straight. A lot goes into walking, different sets of muscle are controlled by the brain. It takes 200 muscles to take a single step forward, maintaining balance takes some brainpower. 

The Moravec’s paradox was articulated by  Hans Moravec, Rodney Brooks, Marvin Minsky and others in the 1980s.

According to Moravec himself, “It is comparatively easy to make computers exhibit adult level performance […] and difficult or impossible to give them the skills of a one-year-old.” 

The Microsoft AI chatbot that got out of control

In March 2016, Microsoft created an AI chatbot called Tay.ai. Tay was created as an experiment in “conversational understanding”. The more Tay chatted with people, the smarter it got. Tay was meant to be an extroverted teen who communicated the way most teens do. Microsoft went on to create a Twitter account for the AI, and learning was underway. 

Things started right with Tay communicating like a nice teen would do. But in less than 24 hours, Tay began to change. She/he (most likely she) started soaking in a wide range of information on Twitter. Soon Tay became an obnoxious robot parrot on Twitter. Tay threw insults at feminists, embraced Adolf Hitler, fell in love with Donald Trump’s racist nature and everything on Twitter that was bad. 

Why Tay only preferred to learn the obnoxious things on the social media site, remains a mystery.

What does AI music sound like?

The entire instrumental and some percentage of the lyrics in  Taryn Southern’s song was created with artificial intelligence. The song was composed with an open-source AI called Amper. All Taryn had to do was to input the genre she wanted, the kind of instruments she wanted, and the AI got to work. The AI is not exactly going to replace Beyoncé or Drake, but we might start seeing AI getting signed to record labels. 

This AI could take my job

An AI programmed to write novels, made it to 7th place out of 1,450 submissions in a national literary prize in Japan. Though AI can write well enough to make it so far in a writing competition, it still needs the help of humans. 

Here’s how the AI-powered novelist wrote the novel, The Day A Computer Wrote a Novel. 

Researchers at the Kimagure Artificial Intelligence Writer project were on a quest to discover the true meaning of creativity. Professor Toshiro Sato a researcher for the project, believed that creating a computer that was capable of creativity would help humans understand the true meaning of our own creativity.  Like humans, AI also needs information to create information. We can never create something out of nothing, the latest art and science are indebted to the ones that preceded them. 

It is for this reason that Sato and his team of researchers fed the AI a set of random sentences and an order of events. In all, the computer did about 20% of the work. That number seems low but Sato explains just how crucial that 20% is. A comparison between writing and the board game known as GO really puts things into perspective. Go is played on a 19×19 board, much more difficult than chess. At the end of the game you would have made about 361 moves at most but with writing, the first word you write is from about a 100,000 choices. 

Language might have a finite set of rules and words the expressions however are infinite. Putting random sentences together in a way that is structured and makes sense is no small feat.

All forms of technology will continue to grow, what excites us today might look ordinary in a few years. AI is growing and will get better.

Comments

Bolu Abiodun is a recent graduate of Theatre and Media Arts, Federal University Oye-Ekiti. A journalist with over a year's experience on the job. A former editor at American Media company Project Forward, he is a skilled content creator, social media manager and digital marketer.

Continue Reading
Click to comment

Leave a comment:

Artificial Intelligence

How AI Is Helping Fintechs Provide Intelligent And Better Financial Services

Published

on

AI fintech services

We live in an era of data. In today’s world, data is the new gold. The quality of services now significantly depends on how much insight can be extracted from data to help in the creation of the services. For fintech organizations, building services that harness the power of data and artificial intelligence has now become necessary to ensure that the services are tailored to meet the needs of customers. Artificial intelligence is now being used in various ways to help fintech companies provide intelligent and improved services. Some of the major areas of AI application in fintech are discussed in this article.

Risk Assessment

From insurance companies to banks and other fintech institutions, assessing credit worthiness and estimating the level of risk associated with every transaction has become very crucial. Now, many fintech companies employ the use of AI in determining the credit profiles of clients which helps to minimize financial losses when customers fail to repay loans or meet other financial commitments. 

Predicting and preventing fraudulent transactions is another challenge that fintechs are using AI to solve. Using machine learning algorithms, fintech organizations are able to build more accurate fraud detection mechanisms to curb the activities of scammers. The advantage of using machine learning for fraud detection in financial systems is that the machine learning model can learn from the financial data by itself. Thus, it is able to uncover hidden patterns and make a more robust prediction compared to traditional fraud detection algorithms. AI-based fraud detection algorithms can also be used to verify insurance claims and flag fraudulent ones. 

Churn Prediction

Customer churn is an important Key Performance Index (KPI) for any organisation. Preventing customer churn is aimpoaaaustomers and improve customer engagement. Many fintechs across the world now use AI to increase customer retention by understanding customer behaviour and making data-driven decisions to retain the audience of customers.

Intelligent Customer Service

Customer service is an aspect of fintech that has been significantly transformed by AI. The use of AI in this area has drastically reduced the need for human customer care representatives and the cost associated with employing these representatives. With AI, more customers can be attended to more efficiently via chatbots, virtual assistants etc. 

Chatbots are, particularly, one of the most common uses of AI in fintech customer service. Chatbots are sophisticated conversational AI applications that can engage with customers, address complaints and basically fill in the gap of a human employee. Chatbots have now become faster and easier means for customers to fix issues they have while using fintech services.

The Future of Fintech With AI

The use of AI in financial technology extends beyond risk assessment, churn prediction and intelligent customer service. Areas like payment processing and sentiment analysis are also being transformed by AI. Organizations like MasterCard and Visa have been able to improve the quality of their services by leveraging AI to achieve this. Personalized banking and financial services will define the future of financial technology. Better experiences will be developed for each customer in a unique and personalized manner. This may be impossible without AI. The future of fintech is geared towards smarter and more intelligent services, with AI steering the wheel to this future.

Comments
Continue Reading

Artificial Intelligence

This New AI Can Tell The Kind Of Faces You Find Attractive

Published

on

AI researchers are constantly coming up with ways to make the technology serve us better. However, while these services are impressive they can be a tad creepy. A time when social media will be able to like pictures on our behalf, based on how our brain reacts to the picture sounds quite uncomfortable. 

Although Instagram isn’t working on an algorithm that tracks brain waves, researchers from the University of Helsinki and Copenhagen University are. This AI matchmaker can tell when you find someone attractive. It creates a database of this information and knows the kind of faces you find attractive or in more lucid terms, “your type”

Here’s how it works

According to Digital Trends, the researchers used a generative adversarial neural network. This network has over 200,000 pictures of celebrities. The AI then recreates original faces based on that database. A group of participants kitted with electroencephalography (EEG) caps were shown these images. Concentrating on which picture they found attractive, the cap read their brain waves to know how they felt about the picture. 

Michiel Spapé, one of the researchers, said, “By capturing the brain waves that occurred just after seeing a face, we estimated whether a face was seen as attractive or not. This information was then used to drive a search within the neural network model — a 512-dimensional ‘face-space’ — and triangulate a point that would match an individual participant’s point of attractivity.”

The AI uses machine learning to connect the dots as per what we are attracted to. Whatever pictures cause a spike in brain waves is stored, the machine analyzes them and looks for the faintest detail they have in common. Afterwards, it comes up with facial features we might not know we are attracted to. 

So, how do you measure attraction?

The researchers have found out that roughly 300 milliseconds after a participant sees an attractive image, the brain lights up. Well, a living brain lights up every time but this particular “lighting up” gives off a specific signal identified as a P300 wave. A P300 wave in itself, does not mean you’re attracted to something. It simply means you have spotted what you have been asked to look out for. The participants have been asked to look out for images they find attractive, as such, any discovery of a P300 wave at that time, means that they find the picture attractive. 

Great algorithm for dating apps?

An algorithm that can tell which facial features you find attractive will be great for dating apps. It will recommend to users, the kind of faces they will probably find attractive. 

However, one can argue that the laws of attraction aren’t that simple. Facial features are just a facet of what people find attractive. Nevertheless, there could be an integration of the detection of all the possible attraction facets. This would go a long way in cementing AI’s position as a major partaker in human lives- or love lives, as the case may be.

Comments
Continue Reading

Artificial Intelligence

Using AI To Grow Avocados And Improve Agricultural Productivity In South Africa

Published

on

Globally, South Africa maintains a spot as one of the major producers of avocado. In 2019, the nation secured approximately 1.1% international market share in the multi-billion dollar avocado export market. However, its annual production volume of between 80,000 and 120,000 tonnes is still many miles behind the annual production volume of 1.5 million tonnes by top avocado producer, Mexico. However, this figure could be significantly increased with precision agriculture, thereby, providing ample opportunity to boost the nation’s overall revenue from avocado export.

The right amount of irrigation is extremely critical for South African avocados to attain maximum growth. For a long time now, South African farmers have relied on traditional irrigation strategies that are largely hinged on their intuition and experience, which has hindered the attainment of the maximum productivity that could be reached in avocado production. Cracking the issue of irrigation could unlock an astonishing increase in South Africa’s annual avocado production volume.

Now, Israeli company, SupPlant, is using technology, particularly AI, to provide farmers with smarter and more efficient ways to improve agricultural productivity. They have been able to achieve this by developing an AI model that employs predictive algorithms to provide customized irrigation recommendations based on analyzing about 100 million data points. Their solution is now being used by South African farmers to grow avocados as the solution is particularly useful for South Africa’s avocado irrigation challenge.

To gather data, SupPlant’s solution involves the use of sensors. These sensors are located in 5 different parts of the plant (avocado, deep soil, shallow soil, leaf, stem/trunk). The sensors monitor plant stress, the exact water content in the soil, health data of the plant, the growth patterns of the plant and the fruit. Climatic data and data on the growth patterns of the plant are also monitored by the sensors. The combination of the data is uploaded to a cloud-based algorithm, at 30-minute intervals, which then makes predictions based on the input data from the sensors to provide farmers with irrigation recommendations.

However, the solution above does not solve all of the challenges that South African farmers face in avocado production. Another critical challenge for the farmers is associated with the weather. To overcome this hurdle for South African farmers, SupPlant integrates the use of top weather intelligence platform, ClimaCell, to monitor the weather in a particular plot of land. Today, with the use of SupPlant’s mobile app, South African farmers can monitor their farm plots and control irrigation on any plot from anywhere. The mobile app provides graphical information on historical and future irrigation plans, present and forecasted climatic data customized for each farm plot, growth patterns of the avocado plant, agronomic insights and recommendations for irrigation.

If, for instance, plant stress begins to increase, farmers will immediately get altered via the mobile app and receive data-driven amplified irrigation recommendations to prevent severe damages to the plant. With SupPlant’s solution in the hands of farmers, farmers will no longer need to engage intuitive and traditional irrigation and farming approaches.

In such a time as now, when the demand to produce more in smarter and more efficient ways is seemingly high, SupPlant’s solution evidently lies beyond South Africa’s avocado industry, as it also extends to other areas where precision agriculture could be harnessed to unlock massive agricultural productivity. With a technology like AI, the journey to smarter and more intelligent farming is just evolving.



















Comments
Continue Reading

TRENDING

%d bloggers like this: