A new survey of 2,000 UK adults by Fountech.ai has revealed the fears people have about artificial intelligence (AI). It found: 

• 61% of people think that AI should not be making any mistakes in its analysis or decision-making 
• The majority (69%) think a human being should always be monitoring and checking decisions that are made by AI 
  • This rises to 77% among over-55s
• 45% say it is harder to forgive errors that are made by machines than it is human mistakes
• Almost three in four (72%) believe that companies that develop AI should be held responsible for any mistakes that the technology makes 
• The majority (64%) want to see more regulation introduced to make AI safer
  • 73% of those aged over 55 shared this view, while only 53% of under 35’s said the same thing

 
People want increased regulation and more accountability in the field of artificial intelligence (AI), new research by Fountech.ai has revealed. 
 
The AI firm commissioned an independent survey among 2,000 UK adults to uncover their attitudes towards the current state of AI development. It found that the majority (64%) want to see more regulation introduced so that the technology is safer to use and does not pose threats to society. 
 
Those aged over 55 appear more sceptical of AI, with almost three quarters (73%) keen to see additional guidelines introduced to improve safety standards. This is in comparison to just over half (53%) of those aged between 18 and 34 who held this view. 
 
The majority of people also hold a high expectation when it comes to AI’s performance and functionality. 61% of respondents said that AI should not be making any mistakes when undertaking an analysis or making decisions. Even more people (69%) think that a human being should always be monitoring and checking decisions that are made by AI. 
 
When questioned about the chances of AI making a miscalculation, 45% of UK adults believe it is harder to forgive mistakes that are made by machines than it is human mistakes. This figure is similar across all demographics. 
 
Elsewhere, Britons also want to see companies take more accountability. 72% of people believe that companies that develop AI should be held responsible for any mistakes that the technology makes. At 81%, those aged over 55 were the most likely to hold this view, while at 60%, millennials were the least likely to agree. 
 
Nikolas Kairinos, founder of Fountech.ai, said: “We are increasingly relying on AI solutions to power decision making, whether that is improving the speed and accuracy of medical diagnoses, or improving road safety through autonomous vehicles. As a non-living entity, people naturally expect AI to function faultlessly, and the results of this research speak for themselves: huge numbers of people want to see enhanced regulation and greater accountability from AI companies. 
 
“It is reasonable for people to harbour concerns about systems that can operate entirely outside human control. AI, like any other modern technology, must be regulated to manage risks and ensure stringent safety standards. That said, the approach to regulation should be a delicate balancing act.
 
“AI must be allowed room to make mistakes and learn from them; it is the only way that this technology will reach new levels of perfection. While lawmakers may need to refine responsibility for AI’s actions as the technology advances, over-regulating AI risks impeding the potential for innovation with AI systems that promise to transform our lives for the better.”

After almost a decade working in a large, global bank, I can speak to the challenges faced by all three lines of defense in trying to combat financial crime. I can also attest to the effect these processes had on our clients.  As a front-line corporate relationship manager, I frequently had to navigate the KYC, remediation and payment screening process for my clients.

Not only was this an incredibly time consuming and frustrating process on an organizational level, but more painful was the deleterious effect it had on our clients and their business: crucial payments to vendors were delayed unnecessarily, accounts took months to open, and required incessant back and forth between multiple parties, and account fundings/transactions always came down to the wire because of basic due diligence, regardless how much work you tried to do ahead of time.

Much of the process that required our intervention seemed mundane, repetitive and inefficient, which compounded everyone’s frustration.

Sound familiar?

These types of repetitive, mundane tasks are ideally suited to be outsourced to artificial intelligence, which the industry seems to now realize.

Artificial intelligence can be an incredibly valuable tool, in that it can offload mundane tasks, provide insight into customer and employee behaviour, create more standardization, and help reduce or manage costs.

But as technology becomes increasingly sophisticated, there are many factors to weigh in the decision making process.

After countless conversations with stakeholders and decision makers in the industry, I have learned that there are 5 main concerns when implementing regulatory technology, especially AI technology, in the financial sector. 

I’ve outlined them below, along with why they matter:

1. How transparent is the AI?
2. What if the AI learns the wrong behaviours, such as bias? 
3. Does it have more than one purpose? What is the roadmap?
4. Is it better than what I have now? More accurate, faster, more standardized, more cost effective? Can ‘better’ be tested quantifiably?
5. What are the redundancies? How will this technology affect my operational resiliency?

Let’s look at each point in order.

How transparent is the AI?

While this seems like a straightforward question, “transparent” really encompasses three separate factors: 

1. Will my team and our stakeholders be able to understand how it works?
2. Will I be able to easily demonstrate to Audit, the Board and regulators that it’s doing what it’s supposed to do?
3. Can I get a snapshot of what is happening at any given moment?

All of the major regulators have stipulated that artificial intelligence solutions be explainable and demonstrable. Both of these concepts are rather self-explanatory but still worth exploring.

Explainability

It’s not sufficient for your compliance team to understand how the AI makes decisions. They also need to be comfortable explaining the process to key stakeholders, whether they are board members, the internal model committee, audit, or the regulators.

If your technical team can’t understand the technology or how decisions are made, or if the vendor claims confidentiality to protect their IP, this is a cause for concern.

Demonstrability

Like transparency, demonstrability captures a few components - it means you have to be able to demonstrate:

• What decisions the AI has made;
• What changes you’ve made to how the AI makes decisions; and
• Who made the changes.

This is where an audit trail comes into play. First of all, is there one? And if so, is it immutable, and does it capture all actions in the AI or just some of them? Is it exportable in report format and, if so, is the report readable and can it be easily understood?

Compliance is a data-driven world, and the risk associated with being deemed non-compliant is substantial. Being able to capture and export changes to, and decisions made within your AI is crucial to your relationships with your stakeholders.

As personal liability expands in the corporate world, board members and committees increasingly require an understanding of not only how compliance risk is being mitigated, but also clear evidence that it’s being done, how and by whom.

What if the AI learns the wrong behaviours, such as bias?

The underlying questions here, without detracting from the very serious concern about embedding existing unconscious bias into your AI, are as follows:

1. If the AI is wrong, or my requirements change, can I fix it? How easily?
2. What impact will tweaking the AI have on everything it’s already learned?

An industry journalist recently asked me if I thought bias was a problem with AI. My answer to her, and to all of you, is that AI simply learns what’s already happening within your organization. As a result, unconscious bias is one of the things that AI can learn, but it doesn’t have to be a problem.

While you can’t really prevent AI from learning from past decisions (that’s kind of the point), good technology should enable you to identify when it’s learned something wrong, and to tweak it easily to prevent bad decision-making from becoming embedded into your AI’s decision making.

This ties in to the need for transparency and reporting.  It’s not only necessary to see how decisions are made, you also need to be able to prevent poor decisions or bias from being part of the AI’s education. And all of these things need to be documented.

When testing out new vendors, once the AI engine has been trained initially for your proof of concept, you should be able to clearly understand the findings, and be able to make changes at that time (and thereafter). You will very likely be surprised by some of the ways decisions are currently being made within your organization.

For example, at Silent Eight, our technology investigates and solves name and transaction alerts for banks. This work is typically done by teams of analysts, who investigate these alerts, and close them as either a true positive (there is risk here) or false positive (there is no risk).  True positive alerts require substantially more time to investigate and close than alerts deemed to be false positive.

Analysts typically have KPIs around the number of alerts they’re expected to investigate and close each week.

By late Friday, the analysts are doing everything they can to make sure they meet this quota. As a result, it’s not unusual during the AI training process that the AI learns that 4pm on Fridays is a great reason to close out pending alerts as false positives.

Obviously this is a good example of AI learning the wrong behaviour and needing to be tweaked. It’s also a good example of mistaking correlation with causation, which is a topic worthy of its own examination on another day.

Today, as regulations are introduced and amended, you’re continually updating your policies to reflect these changes. It’s no different with artificial intelligence.  It’s imperative that your AI engine is correspondingly easy to tweak, and that when you tweak it, you don’t lose everything it has already learned.

Thoughtful, well architected technology should be built in a manner that makes it easy to update or amend part of the AI engine, without impacting the rest of the learnings.  This is something you should both ask about, and test in the POV environment.

 Does the AI have more than one purpose? What is the roadmap?

Many financial institutions have the duelling mandates to be both innovative and transform digitally, but also rationalize vendors. So, when considering artificial intelligence solutions, which are often niche, it’s worthwhile finding out:

• How the vendors decide to build out features;
• Whether they are willing to customize their offering for you;
• How reliably they’ve delivered on features in the past; and
• Whether what’s on their roadmap adds value for you.

This way you can ensure that the decision you’re making is one that is future-proofed and set up for longevity.

Is it better than what you have now?

Better can mean different things to different organizations and individuals. It’s typically tied into the problems you’re experiencing now, and what your organization’s strategic focus and priorities are.  When I ask clients and prospects what they mean by ‘better’ the answers I hear most commonly are:

• Is it more accurate?
• Is it faster?
• Will it give me greater standardization?
• Will it enable me to identify more risk?
• Will it enable me to federate by jurisdiction?
• Will it lead to greater efficiencies?
• Is it more cost effective?
• Does it increase my visibility? Ie. is it transparent?

Once you’ve defined what ‘better’ means to you and your organization, you need to find out from your prospective vendors if and how ‘better’ can be tested quantifiably.

What are the third party dependencies? How will this technology affect my operational resiliency?

Operational resiliency and third party due diligence have become a significant focus in the industry and can be a barrier to doing business.  Many regulators, including the EBA and the FCA, have issued guidelines on the topic, and continue to revisit it

It’s vital to understand if a vendor is reliant on any other vendors in their tech stack, if they’re using open source code, what their deployment is (on premise, in the cloud, in a private cloud), and what security standards they adhere to.

Take back the things you can control

Right now, the financial services industry is beset by many challenges that are outside of its control, including the low interest environment, working remotely, bad debt provisions, and the increased new accounts and suspicious activity resulting from COVID.

Your compliance costs and processes are a piece of the puzzle you can control.  Good artificial intelligence technology will enable you to offload some of your mundane, repetitive tasks, freeing up you and your team to focus on more complex risks and higher value projects.

I recognize that artificial intelligence can be a bit daunting, and that it has a mixed reputation in the industry.  However, if you’re armed with a dose of skepticism, the right questions to ask, and you approach it with an open mind, you’ll be amazed by what it can do.

Amber Sutherland has over fifteen years experience in finance, business development, and regulatory technology. She worked for a large global bank for a decade, and has experienced first hand the evolution banks go through in their digital transformation and compliance journeys. This experience led her to move to regulatory technology, where she helped a former client enter the UK market. She then joined Silent Eight to help enter the EMEA market and grow their business in EMEA and APAC. Silent Eight is an AI-based name, entity and transaction adjudication solution provider to financial institutions.

Amber’s passionate about finance, technology and trying to prevent criminals from benefiting from their crimes. In her spare time, you will find her on the ski hill. She is currently SVP of Sales, EMEA & APAC for Silent Eight. Follow her on Linked in. linkedin.com/in/amber-sutherland.

You may rely on reviews to be a smart consumer and make a good choice for a purchase, but it’s possible that you’re reading an opinion generated by a machine. The use of AI in online reviews searches for and compiles reviews from multiple sources. In addition, AI has also been employed to write reviews, which makes online shopping even more complex than it already was.

Reviews Matter

Since reviews of products or services can make or break a business, they take on great importance in today’s world of consumerism. Enough bad reviews can scare away new and old customers, while conversely, glowing 5-star reviews can attract and drive more sales.

Savvy companies keep a close eye on individual reviews, trends in critiques, and the average star rating for their products. However, with such a prolific selection of online retailers, it can be hard to track all the places a product may be reviewed. More importantly, social media reviews don’t offer a star rating system but rather consist only of text comments and perhaps an emoji or two. For statisticians, that makes the waters very muddy.

The AI Advantage

Searching across all the posts is where AI makes a difference. There are a number of AI-driven services that scour social media platforms and retailers, big and small, for mentions of a product, brand, or company, gather up all of the reviews relevant to the search criteria, and analyze the written reviews to pick out the trends and common themes that may be invisible to someone looking at star ratings or just the company’s customer complaint logs. These systems can go from basic scraping of data for keywords to full analysis of reviews for negative tones, new ways of using a product, or demographics of users that like and dislike a service.

For the past few years, AI has been able to and employed to write reviews. Reviews that lead you to buy a product, go to a restaurant, or enlist a service provider may be written by an AI program. This can allow an unscrupulous supplier to turn on the bot and have it churn out reviews praising their own product on top retailer sites, their own site, and on social media, posing as real users posting believable reviews. While unethical and misleading, this type of behavior would be difficult to track down if done right, and with AI getting better at impersonating human typed speech every day, may not even set off alarm bells to have anyone look into the behavior.

On the flip-side of this nefarious use of AI, companies can enlist an AI reviewer to tirelessly spit out convincing reviews of competitive products, highlighting shortcomings, poor quality, lack of service, or anything that will drive sales down. The target company may respond to these reviews, show consumer good will, and try to reach out to smooth the waters, but a string of 1 and 2 star reviews may create enough damage to spell doom for a product or company.

How To Use Reviews

As AI further infiltrates online consumer reviews, it will be up to the consumer to decide which reviews to trust and which to question. Users can no longer believe everything they read about a product or service, but it’s a good start. Look for common denominators between the reviews, and while you might not spot the computerized versions, you could still use reviews to inform your choices.

Artificial Intelligence (AI) has achieved a lot since it came into existence. However, the main objective has always been to get high-level AI systems that mimic humans in most of the things they do. In essence, the primary objective of AI systems is to function independently, like humans. However, that AI is conscious is no longer a matter of if but when. This was envisioned by Norbert Wiener, the renowned father of cybernetics in his book The Human Use of Human Beings, back in 1950. Although Norbert lived at a time when vacuum tubes were the building blocks of computers that were few at a time, he foresaw how automation would change the society for the better.

Unlike other researchers who see AI as a technology that will imitate and replace humans, Norbert noted that this technology would change humans in many ways, including creative activities. Comparing the technology with a whirlpool, he noted that we would not abide by what humans have been doing for ages, but we will instead conform to the new changes.

Here we are in 2020 where artificial intelligence is the talk of the day. With the focus increasing on the potential use and misuse of this technology, the question that many people are now asking is whether AI can ever be considered to be fully conscious or not. The majority of experts believe that we are decades away from seeing AI emerging as conscious. Although many notable achievements have been made in its research, no one fully understands what conscious AI looks like. As deep learning continues advancing, many researchers now believe that the time for machines to think like humans is not so far like it was in the 1950s or even 2000s. Others, however, believe that we will need to do more with neural networks and algorithms, while others still believe that machines will never match humans at all.

Although many scientists may have a better understanding of the AI systems, they have little knowledge of intelligence in relation to the human brain or consciousness. This means that we do not fully understand what intelligence is when it comes to humans, that can make it easy to compare human to AI systems. Going by this, what we know is some neural networks and algorithms but little understanding of whether AI has or will ever reach the level of human intelligence. It is also hard to demonstrate a path towards AI consciousness, although it is not impossible.

According to Johannes Kleiner and Sean Tull, AI might already be conscious. This is if their research paper on the nature of consciousness is anything to go by. According to their research, the duo notes that the universe and anything in it have some form of consciousness. They go ahead to note that anything that exists in the universe exhibits some characteristics of some degree of consciousness. Their argument suggests that consciousness emerges as a result of physical states and that human consciousness is due to the ability to experience things. For instance, a tree is conscious because it has the ability to sense sunlight and may even bend towards it. Since a tree is a living thing, it is hard to draw a comparison between them with the AI systems.

While many theories are already there on consciousness, it is hard to point out precisely whether AI is already conscious or we have to wait longer before AI can mimic humans not only in their actions but in thinking as well. One thing we can all agree on despite varying opinions and views is that AI has advanced to the extent that it can now compete with humans in many areas. More is expected in the coming years as new inventions keep coming.

According to ZD Net, business leaders and researchers working with artificial intelligence have widely different ideas of how it should be used. 

Business leaders and technologists see artificial intelligence differently -- which is not a surprise. 

Read the article on ZD Net

According to ED Hat, artificial intelligence can be used to localize predictions on the spread of Covid-19.

Despite efforts throughout the United States last spring to suppress the spread of the novel coronavirus, states across the country have experienced spikes in the past several weeks. 

Read the article on ED Hat

While other industries have found a way to move forward during Covid-19, the artificial intelligence industry is still struggling to find its way, reports Tech Crunch. 

2020 has made every industry reimagine how to move forward in light of COVID-19: civil rights movements, an election year and countless other big news moments.

Read the article on Tech Crunch

Technologies like artificial intelligence and machine learning have transformed business processes across different industries, changing how companies operate, manage finances, gather and analyze data, and build customer relations.

As the algorithms and data-driven methods continue to develop, venture capitalists are becoming more interested to invest in artificial intelligence startups providing the most beneficial solutions. According to data gathered by BuyShares.co.nz, the total amount of funds AI startups raised over time reached $61.6bn in the second quarter of 2020, a 35% increase in a year.

More than $8bn Funds Raised in 2020

In 2010, artificial intelligence startups worldwide raised a total of $277.3 million in funding rounds, revealed the CrunchBase data. In the next three years, this figure jumped to over $1bn for the first time.

With the increasing number of investments in companies providing artificial intelligence solutions, the total funding amount reached $26.6bn in the second quarter of 2018, a tenfold increase in eight years. Statistics show that 2019 has witnessed the most significant rise in the value of investments, with the total figure growing by $19.1bn and reaching $53.6bn at the end of the year.

Increased investments into artificial intelligence companies continued in 2020 with the total amount of raised funds growing by $8.2bn from January to July, despite the slowdown in venture capital funding caused by the COVID-19 pandemic.

In February 2020, the Chinese autonomous vehicle startup Pony.ai announced it had raised $400 million in a funding round from Toyota Motor Corporation, the most significant investment in 2020. Last year, the two companies allied in testing autonomous vehicles on Chinese public roads using Toyota’s Lexus vehicles piloted by Pony.ai’s self-driving system. The company will use the new funds to deepen the collaboration with Toyota on self-driving technological development, while making a push into mobility services in China.

In March, Chinese big data company MiningLamp raised $300 million in a round led By Tencent and Temasek, the second-largest investment this year.

Statistics show that Berkshire Grey `s $263 million worth funding round led by SoftBank Group represents the third-largest investment in 2020. The Massachusetts-based robotics and AI company delivering retail, eCommerce, and logistics fulfillment automation, raised funds to finance global expansion, acquisitions, and team growth.

North American AI Startups Raised $33.7bn

Analyzed by geography, North America represents the leading region with a total of $33.7bn worth investments in artificial intelligence startups. The US companies raised more than 95% of that amount, with California, San Francisco, and New York as the main hubs.

Statistics show Asian companies raised a total of $20.3bn in funding, ranking as the second-leading region globally. European AI startups follow with $7.3bn worth investments so far.

The CrunchBase data also revealed that the five most active investors in this industry come from the United States. The Colorado-based global platform for investment and innovation, Techstars, ranked first on this list with 258 funding rounds so far. Statistics show that startup accelerator Y Combinator, the second most active investor, took part in 198 funding rounds. Another US innovation platform, Plug and Play, follows with 104 funding rounds. Venture capitals firms SOSV and 500 Startups ranked fourth and fifth on this list, with 99 and 83 funding rounds, respectively.

The full story can be read here: https://buyshares.co.nz/2020/07/21/artificial-intelligence-startups-raised-61-6bn-in-total-funding-a-35-jump-in-a-year/

Like several other banks in the EU reeling from the impact of the 2009 financial crisis, our client Bank (name withheld for confidentiality) had to take measures to restructure the non-performing loans under a parliament-approved securitization scheme. The Bank uses its in-house call center to contact non-performing loan debtors and negotiate offers to restructure the debt according to the debtor’s unique situation and their ability to pay-out the loan.

Given the effect of non-performing loans (NPLs) on its bottom line, the Bank was seeking a technology partner to help it optimize the collections process. Behavioral Signals partnered with this Bank and applied its AI-based agent-customer matching technology towards maximizing the effectiveness of their Call Center. 

AI and Finance

Financial institutions are leveraging AI systems in their contact centers for improved efficiency in delivering service to customers, helping them engage the customer uniquely. The ability to route and measure agent-customer interaction offers not only a higher probability of achieving the desired result, be it a sale or collection, but also insights on how AI is impacting the whole communication process.

Behavioral Signals  proposed a model to optimize the call centers’ outcomes by using machine learning techniques to route the  caller to the best-matching agent. By creating agent and customer behavioral profiles, from their past audio interactions, we were able to create a predictive model on which agents’ skills and overall behavioral traits during the calls are best suited for the specific customer in order to achieve the desired result.

From Voice to Revenue

The solution was deployed on Live-Data in a champion/challenger setup demonstrating a significant return-on-investment with an actual increase of active debt restructuring applications by 20.1% in total. On top of that, this improvement was essentially achieved with 7.6% fewer calls (Right Party Contacts), leading to additional cost reductions. In absolute numbers, these results actually correspond to approximately $7.5M USD of additional restructured debt for the Bank over the course of this evaluation and a potential $300M USD annual upside.

Next-Gen Banking

Contrary to a common concern, this drastic improvement of collection outcomes did not compromise the customer’s satisfaction levels. The Bank noted that these interactions have been smoother than average with the majority of interactions rated very high by the Bank’s QA team, clearly underscoring an improved ability of the agents to build rapport with their clients. Furthermore, these interactions provide excellent examples of successfully-handled interactions that can be used in training/ coaching of new and existing agents, thus improving the performance of the call center as a whole.

The Bank is extending the scope of cooperation with Behavioral Signals to deploy the Behavioral Profile Pairing (agent customer matching) application to the entire debt collection team along with other applications, to improve the efficiency and performance of its call center. 

It has now become common hearing people talking about how artificial intelligence (AI) will change various aspects and industries in every country on the planet. There are growing fears that this technology will eliminate many jobs and professions through automation of tasks. Many reports have shown the truth behind job automation especially for certain professions and demographics. For example, a report by Brookings Institution noted that automation threatens more than 25% of jobs in the US, most done by low-wage earners where tasks are repetitive and routine-based.

Despite these harsh realities, we have no option but to conform to the new changes and adopt these technologies due to the massive opportunities presented that far outweigh downsides. Although artificial intelligence will indeed make other tasks obsolete, it will also inject more jobs and change the existing ones for the better. Workers in different countries are now beginning to see the bright side of AI on their roles, and organizations and are slowly starting to change their minds on this technology. Most of them are now grateful for having robots as their coworkers because of how they simplify and streamline their lives.   

A majority of human workers now trust orders given by robots over those of human managers. Most of them are even turning to robots for advice instead of their human managers. In some companies like American Express, for example, tasks such as identifying product offers that are relevant to a specific customer segment are now made by AI, therefore taking over the role of managers and employees in handling these tasks.

AI is now taking over administrative tasks traditionally done by human managers while the roles of managers are now evolving as they focus more on soft skills over hard ones. Surveys have indicated that robots are far much better than human managers in doing tasks that require hard skills as they can easily give unbiased information, maintain work schedule, solve problems quickly and their budgets can also be easily managed. On the other hand, humans can complement them with better empathy. Unlike humans, robots can also provide answers to confidential information without causing fear among the interviewees. They can also evaluate performance of the team with no bias. Although robots and AI systems have advanced significantly, human managers still beat them in many areas. For instance, human managers can understand feelings of other humans better and can reliably provide oversight and direction.

Artificial intelligence collaboration with humans in workplaces is freeing up managers, giving them time to focus on other issues of their jobs. By cutting managers loose from their traditional tasks, AI is enhancing innovation. They allow managers to have more impact on their work by giving them time to create strong relationships with other employees that in the end impacts their roles.

Far from replacing managers, AI is improving their decision-making capabilities. The main change in decision making in management originates from the AI’s deep learning abilities. Machines through deep learning are becoming smarter with time as they continually learn from the experience of human managers just like the way managers gain their knowledge from interacting with other managers and encountering new challenges. With deep learning, AI promises to be part of decision making process in the coming days. Through the help of AI, managers can tap into the advice they can offer at any time due to large data they have. This will make managers smarter and more qualified to make informed decisions that can take a company to the next level.

From Super Bowl ads boasting the benefits of machine learning to rappers touting the advantages of artificial intelligence, nearly everyone knows about the cloud, how powerful it is, and the advantages it can provide for both businesses and consumers. Then you also have the pundits adding to the hype with absolute statements like this:

“If you have not developed a cloud-first strategy yet, you are likely falling behind your competitors” -Gartner

If you can find a flaw with that statement, then you’re ahead of the game.

Cloud-first is a brilliant strategy that has served as the de facto CIO direction for close to a decade. Nearly every CIO serving entities from the government to leading enterprises publicly embrace this approach. The cloud-first strategy has resulted in a predestined design where cloud-compatible data and applications find their way to the cloud to achieve efficiency and flexibility.

Cloud technologies are so powerful and beneficial that once an organizational cloud standard is in place, it is difficult to see favor in any other information systems. In so many ways, it has never been simpler to get on a cloud platform. One credit card swipe, and you’re on your way. As a result of these conveniences and benefits, we live in a world where public clouds, private clouds, cloud services, and a litany of cloud technologies help improve business power and add to the flexibility of organizations.  

A Cloud Wake-Up Call

Despite all the advantages of cloud computing, many organizations under a cloud mandate today find their technology, operating, and migration costs are not what they anticipated. Many IT departments have legacy applications that are not cloud-ready, plus significant shadow IT and overly complex cloud infrastructures are creating severe cloud sprawl. Other organizations can't build technology around their business initiatives because they don’t have the staff, or the necessary efforts are beyond their capabilities. For these and other cloud situations, a cloud-first approach is making matters worse.

For these and other situations, the cloud-first strategy has uncovered a whole new world of challenges: 

• The need for cost and consumption control tools 
• Complex compliance mandates and technical requirements
• Cloud staff acquisition and retention
• Increased security needs
• Lack of visibility and metrics
• Bifurcation of IT infrastructures
• Underperformance of legacy applications 

A New Cloud Phase

Hybrid and multi-cloud technologies have proven their effectiveness in the enterprise as the best path forward by adding functional co-existence to the informational technology picture and unlocking the potential for an application-centric cloud strategy. Cloud-leaning philosophies hold untold business values, but not every application out there was built with cloud technologies in mind. There are far more traditionally structured applications than not. Some applications might be better suited to one cloud service over a given cloud-based solution. Factors like security, integrity, auditing, and performance might become more complicated or costly within the constructs of a limitless cloud environment. Depending on your organization’s capabilities, some applications are simpler to leave and integrate through on-premise systems.

Application First

In this second decade of cloud, a new phase is emerging in which business advantages must be achieved by focusing on the user and taking an “application first” strategy rather than a cloud-first strategy. While public cloud and private cloud architectures have extremely powerful benefits, there are too many challenges to navigate when enterprises insist on solely implementing one versus the other. As part of a strategic architectural practice, hybrid cloud and multi-cloud approaches can properly shift the focus to the users and the type of applications they interact with.  

Being smart about the cloud means using its benefits correctly, according to your business, your users, and your applications. We can even go so far as to say that if you have not developed an application and user-first strategy as of 2020, you will fall behind your competitors. Once you’ve built an intelligent strategy, it’s time to move on to a discussion of platform and infrastructure advantages, such as edge computing, artificial intelligence, machine learning, hybrid, auto scaling—technologies that are only possible with the more flexible application-first approach within hybrid, multi-cloud solutions, leaving cloud-first as an antiquated ideology.

Emil Sayegh is widely recognized as one of the industry’s cloud visionaries and "fathers of OpenStack" (he launched and led the cloud computing and hosting businesses for HP and Rackspace), can talk about everything from rising decentralization, to hybrid and edge clouds, and the significant impact that AI and Blockchain technologies will have on the industry. 

Artificial Intelligence is slowly taking over our day to day lives - especially when it comes to the workplace and using tech to enhance and fast-track tasks. There are many sectors in which AI can help improve services - specifically national health services, banking and legal services. Not only do some of the biggest names in tech debate over the future of AI and how it continues to make life easier and better for many, but we are now seeing some of the biggest tech companies now recruiting heavily in AI.

But, who is recruiting for AI roles the most and is making big steps in looking toward the future?

With AI-related jobs more than doubling over the past three years and job postings related to AI increasing by 119%, RS Components has analysed job posts from some of the world’s biggest tech companies to discover who has the highest percentage of AI-related job openings.

Microsoft leads the way with 1,964 AI job openings, equating to 36% of the total jobs they have on offer. With Bill Gates being one of the world’s richest people still, this is no surprise - Microsoft still leads the way with some of the most innovative and thought-provoking technological inventions.

Nvidia follows in second place with 837 of AI job openings, equating to 33% of total jobs they have on offer. The computer game company designs graphic processing units for the gaming and professional markets, so using systems and analytics such as AI helps with processes and development of new games.

A former Google Exec has predicted that AI will replace 40 per cent of jobs in the next 15 years. The Google Exec stated that AI will allow entire new industries to be built from the ground up with automation in mind. AI will be cheaper than human employees, however, it will also create new roles, as well as replace current ones.

Popular social media platform Facebook and streaming platform Netflix make it into the top ten with Facebook offering 191 jobs in AI, 6% of the total, and Netflix offering 35, also making up 6% of the total jobs they have on offer. Both companies are growing rapidly and are having to set new trends to keep up with technology, so it's only right that they are in the top 10 companies that are hiring in AI the most.

The largest tech company in the world, Apple, comes in at number 12 with 129 AI jobs on offer, only 3% of their total 4,574 jobs available. Apple is one of the most innovative companies in the world - alongside Google - so it may come as a surprise that they are in 12th place. This could be due to the fact that they are forward-thinking and hired the majority of their AI roles a few years back.

Another innovative company in the technology business is Bezos' e-commerce platform, Amazon, which comes in 15th place offering 297 roles in AI, 2% of its total 18,781 jobs on offer.

It’s clear that the technology sector prioritises hiring in AI, but which sectors will be next to drive up the market and start boosting their recruitment in AI?

Take a look at RS Components’ interactive campaign here which highlights all of the companies which are recruiting in AI.

Top Ten Global Companies with highest % of AI Jobs:

 Bias is one of the biggest documented weaknesses of humans. This can be seen from implicit tests that show that we are not even aware of some of our biases. Although we may say that these biases are part of human nature, experiments have shown that they can affect our judgments and outcomes. Over the past few years, there has been an argument on whether human biases can be solved by artificial intelligence or whether these biases can make their way into AI systems and cause harm. This comes at a time when AI is being deployed by many companies across different industries. The industries need to be aware of the risks of biases ad how AI can be affected by it.

All AI models are developed by human beings and can, without a doubt, have some form of bias. For instance, machine learning algorithms have biases that are a reflection of organizational teams, designer teams, and data scientists. They are also vulnerable to the prejudices of data engineers who gather data. Just like the lack of fairness that we have been experiencing from human decision-makers in different areas, AI is expected to also suffer from it as these prejudices can find their way into algorithms in a variety of ways.

AI systems learn to make decisions based on training data that can be as a result of human prejudice. As such, decisions that these systems can arrive at may sometimes be biased. They can be seen in areas such as gender, race, and sexual orientation, among others. Amazon, for instance, stopped using its hiring algorithms after it emerged that it favored applicants based on specific words such as “execute” or “captured” that were common in men’s resumes. Another major flaw in AI systems that results in bias is data sampling whereby groups are over or underrepresented in the training data. This results in specific groups, such as minorities and women suffering.

Despite the possibility of bias, AI will have some level of trustworthiness, which is a backbone of machine learning. However, even a trustworthy model will have some form of weakness that must be looked into if fair systems are to ve developed. It is upon humans to ensure that training data is fair and representative by all standards. To achieve this and ensure optimal results, organizations must have tech teams with diverse members. These members should be in charge of building AI models and creating training data to be used by AI systems.  Furthermore, organizations should try to find out a highly comprehensive data and experiment it with different metrics and datasets.

As machine learning and artificial intelligence systems increasingly become complex, it has become critical now than ever to have training data that is annotated by humans in an unbiased manner. During the training of machine learning models, human bias can damage the accuracy of these models. If there is are in-house teams that are involved in interpreting training data, they should adhere to an unbiased approach in the classification of data. This may involve using different approaches and styles that are diverse.

Without a diverse approach in classifying data, there is a high risk of less accurate models being created. On the other hand, the external partners who are involved in the collection or processing of data should ensure the job is done by diverse crowds to ensure data is representative. The data annotation design tasks should be carried out carefully. Once the training data has been created, it should be checked to identify whether data has any bias or not.

Minecraft is a popular sandbox multiplayer online building game. It starts with a generated world where terrain and features are created with building blocks. There are also generated characters called “mobs” that appear throughout the world. The two types of gameplay in Minecraft are survival mode and creative mode. In survival mode the player can be harmed and has limited resources. In creative mode the player does not have to worry about being harmed and has access to all resources. In creative mode the player’s imagination is the limit and there have been endless creative projects that have been carried out in the game including recreations of cities and landmarks among others. 

CraftAssist is a framework created in 2019 that features dialogue-enabled interactive assistants. These assistants collaborate with human players in the game. The assistants can build, move objects, spawn mobs and communicate with players by chat. The framework is intended to research how agents interact with players and discover the possibilities between them. The framework is open source to encourage AI researchers to use the toolset. 

Players can interact with other players and the AI assistants, just like in a regular game of Minecraft. CraftAssist includes a baseline assistant, who brings over intelligence of previous interactions. There is also a house dataset that was put together using crowd sourced participants. These participants were tasked with constructing a house within 30 minutes and their actions were recorded. Both of these datasets are part of the core CraftAssist release.

Minecraft is played through a client/server architecture and the bot communicates through the Minecraft protocol. The Minecraft protocol does not require server side mods, which allows the bot to join in with other players and bots. The bot assistant is modular and this allows research to be conducted based on the individual models. These include modules related to perception and memory among others. The bot modules work together to perform tasks and interact with the human players. A human player may request that a bot build part of a house such as the roof. The bot then takes the chat commands and performs high-level actions that carry out the informal chat instructions to build the roof. 

Another example would be a player telling the AI assistant to build a house in the shape of a rectangle near the water. The modules take the given instructions from chat and create a program for the assistant. In this example the object water would be identified as being close and the assistant would be given coordinates to move to and start building. Machine learning works well for difficult and well defined tasks. One of the larger goals with CraftAssist is a framework that takes the often ambiguous human language commands and creates executable actions through the assistant. The interaction between humans and bots is a core focus of CraftAssist. 

More data from past interactions is being added to the assistant for further improvement. If you are a Minecraft player I would recommend giving CraftAssist a try to help build and improve the CraftAssist project!

CraftAssist Documentation:

https://research.fb.com/wp-content/uploads/2019/07/CraftAssist-A-Framework-for-Dialogue-enabled-Interactive-Agents-v3.pdf?

https://github.com/facebookresearch/craftassist

For decades, artificial intelligence has been seeking to mimic humans and nature. The main goal here is to create systems that have the ability to make decisions like humans. As such, most AI systems are brain-inspired, and scientists behind it have always attempted to mimic neural networks, a term borrowed from biology. With computers now able to think like humans, computer scientists are now focused on a different goal as they seek to revise artificial intelligence by transitioning it into smarter and more efficient systems. Initially, most of the research was devoted to training AI through machine learning to enable algorithms to learn how to recognize objects and how to perform basic tasks. Now that this has been achieved, these techniques are being deployed to create more advanced systems that can work autonomously without human intervention.

Google has been working on AutoML (automatic machine learning system), a smart machine learning system that can be mutated into a next-generation AI. This new system has been enhanced through the addition of the Darwinian evolution concept that allows building of AI products that can improve themselves. The product of this improvement is called AutoML-Zero. While this system can be worrying to many, if used well, it can open doors for the development of much smarter systems that can mimic human brains in development and decision making through enhanced evolution.

Going back to Charles Darwin’s concept of evolution, slight or random changes in the genetic makeup in an organism are either advantageous or disadvantageous. If the mutation allows the organism to survive and reproduce, the characteristics are then inherited by the future generation. Otherwise, the mutation dies with the organism. In AI, this phenomenon is known as neuro-evolution. It is a process in that neural networks recreates that built parts of the brain to allow only the strong or smart to survive.

An AutoML system makes it easier for applications to use machine learning and other automated features, unlike AutoML-Zero, that requires human input. It uses a three-step process that is setup, predict and learn to improve itself. The system begins with a selection of 100 algorithms by analysis of their mathematical operations. It then identifies the best of these algorithms through a trial-and-error approach. The best ones are retained and tweaked for the next round of trial-and-error while those found to be less effective are dropped. This is basically a mutation of neural networks. The process goes on and on until the neural network is highly efficient.

The new code is more effective and can efficiently identify things that have always been hard for AI to differentiate, such as the difference between objects. These best performing algorithms are kept in a similar manner as survival for the fittest fronted by Charles Darwin. According to researchers at Google, this Darwinian concept of neuroevolution can lead to the creation of an algorithm with 94% accuracy. Various studies saw an improvement in the algorithm’s image recognition skills.

The evolution of the machine learning algorithms, according to researchers, has led to the creation of more artificial intelligence systems that can work more like humans and without bias, that humans are known for. The mutation of algorithms will see the emergence of systems that can perform several tasks, including complex tasks that have for long relied on human expertise.

Despite the complex nature of the development of AI systems and advanced research that is needed to deliver highly efficient systems, Google’s AI mutation approach has shown to be a giant step towards making cutting-edge AI systems that can be helpful in the future. It is through such initiatives that we are likely to see systems that can outsmart humans in every aspect sooner rather than later.

While artificial intelligence (AI) continues making headlines and breaking barriers that were thought to be impossible a decade ago, humans are now looking at it with both awe and fear. This is understandable because no one can allow their work to be taken away or to be replaced in places where they have for years looked for their daily bread. When people hear that companies are building AI-powered assistants to help in different tasks, they get afraid that AI is close to becoming independent of humans. The questions that they often ask are; Is AI close to matching human intelligence? Are we close to the fictional Skynet predicted in the Terminator? Is AI smart in its own and will it take positions long occupied by humans?

Until a few years ago, humans seem invincible, and computers appeared only servants who cannot outshine the master. In the mastery of language, for example, computers were far second to humans in producing sentences that make sense. However, as the fields of machine learning and natural language processing (NLP) continued advancing, signs were written all over showing the capabilities that computers had in this area. The advances that have been made in NLP has seen computers getting closer and closer to humans as they can now generate complex passages with a click of a button. These advances are made possible by deep-learning algorithms that can pick statistical patterns in words used by humans and the structure of words from many lines of text. This is a step in the right direction. However, despite the worries of humans and the excitement by professionals that AI is close to matching humans in their language capacity, the truth is that computers are not there yet.

According to new writing done by Allen Institute of Artificial Intelligence, there is something that machines still lack, but humans have. Machines do not understand what they are writing or reading. This is the biggest flaw that is yet to be sorted out in pursuit of generalizable AI. Although this question applies to academia more, it also touches on the consumers of AI technologies. As organizations continue to adopt chatbots and voice assistants to help in their daily tasks, lack of adequate understanding of language by these systems is proving to be a  hurdle. This is because it slows access to crucial services mainly in financial institutions and makes these applications prone to failure.

Based on the Winograd Schema Challenge built in 2011 to evaluate the reasoning of NLP systems, it emerged that computers have a poor grasp of semantics and still needs training before they can match humans. The challenge is made up of 273 questions that have pairs of sentences that are identical except for one word known as a trigger. For a system to be regarded as fully intelligence, it must figure out the difference in two different sentences.

Although AI is progressing rapidly, its capability to pass the Winograd Schema Challenge or Turing test, that is designed to see if machines can mimic humans in everything successfully will still need help from humans.  Human touch is still essential for machines to attain full independence, and this will be the case for some years to come. Without help, every word that machines utter will have some type of emptiness and will always be robotic. Humans still have unbelievable levels of accuracy, that stand at 94% compared computers that have between 59% and 79% accuracy. This data will serve as a new benchmark and will inspire further research beyond machine learning.

Unlike humans who can have knowledge of two languages and can mix languages and still make sense, machines have some weaknesses doing so. Furthermore, humans can understand questions easily and choose the correct one accurately. With increased training and improvement in both hardware and software, AI systems will get closer to humans in future.