✕ סגור 
צור קשר
תודה על ההתעניינות .

Thank you! Your submission has been received!

Oops! Something went wrong while submitting the form

impact-artificial-intelligence-vicky-falconer

Vicky Falconer
|
קלה
|
Sep 23, 2016
alt="facebook"alt="linkedin"להרשמה לניוזלטר

Just recently I was asked the question “does Oracle have an artificial intelligence product; I have a customer who is interested in exploring artificial intelligence further…” This prompts the question “interested in what, and why?” Which got me thinking … what do we mean when we say “artificial intelligence” and is there any such thing as an AI product?

Artificial Intelligence (AI) has been part of the imagined future of Science Fiction since the inception of the genre; I certainly became a fan as I delved into the imaginations of Asimov, Clarke, Stephenson, Banks and many others as a teenager.

 Artificial Intelligence is the theory and development of computer systems able to perform tasks normally requiring human intelligence, such as visual perception, speech recognition, decision-making, machine learning and translation between languages. At this point I’d like to highlight that we are not talking about thinking machines in the sense of machines with consciousness or self-awareness, merely machines that perform human-like tasks. There is also the important distinction between performing tasks and “understanding” those tasks. The other important definition is that of Machine Learning – often referred to as a subset of AI, machine learning uses statistics to develop self learning models.

In the AI field there are also two important distinctions – Narrow AI and General AI. Narrow AI refers to machines capable of completing specific narrow tasks – driving a car, analyzing an image, recognising speech, making a movie recommendation, estimating the probability of a legal case being successful and it is in this area that we are seeing so much progress. General AI refers to machines with the ability to apply intelligence to any problem, rather than just one specific problem. There is a fair amount of controversy as to how far away we are from General AI so for now we will focus on Narrow AI.

Both Oracle and SalesForce have made announcements this week of building Narrow AI functionality into customer applications. Oracle’s Adaptive Intelligence Applications are based on insights gleaned from Oracle’s Data Cloud, using anonymised web scale data and applying data science to target business outcomes, including next best offer, best-fit candidates and optimised payment terms.  The system is continuously learning and adapting as the end user interacts with the system. This highlights that the tipping point has been reached – products with embedded AI (such as Clinical Decision Support), AI as the basis for new products (like Siri or Quill) and AI-as-a-Service have all come of age.

So what are the implications for businesses? In a world where more and more products have embedded AI it can be a winner takes all! If your algorithms are better at attracting and retaining customers then you collect more data which improves the algorithms –the network effects are self re-enforcing! It’s not too hard to envisage a world where we have many brands of driverless cars on the roads but the AI or “brain” in each car comes from a single organisation, and that single organisation is collecting all the data from every driverless car to further improve the AI. It’s hard to be the number two competitor in a market where more data means better safety. To this point there are sure to be government policy initiatives to encourage or mandate the sharing of data from driverless vehicles (a topic for another day).

Embedding AI functionality into existing products provides differentiation that is difficult or near impossible for competitors to re-produce. Suddenly we move from feature/function differentiation to differentiation based on the quality of the data and the AI. Highly relevant in the software industry as this week’s announcements confirm.

So what should companies be thinking about? What are the opportunities for embedding AI into your products or processes? Do you want to be the consumer of AI from third parties or a creator of AI for your industry?  AI and big data are two clearly linked trends and for most of the customers I talk to it will be a natural transition from big data initiatives to AI initiatives, leveraging the investment in big data. The other interesting trend is AI- as-a-service, for those customers not interested in building a capability from the ground up. It is certainly time to be assessing the impact of AI on your business processes, business model and importantly on the products or services you produce. 

Just recently I was asked the question “does Oracle have an artificial intelligence product; I have a customer who is interested in exploring artificial intelligence further…” This prompts the question “interested in what, and why?” Which got me thinking … what do we mean when we say “artificial intelligence” and is there any such thing as an AI product?

Artificial Intelligence (AI) has been part of the imagined future of Science Fiction since the inception of the genre; I certainly became a fan as I delved into the imaginations of Asimov, Clarke, Stephenson, Banks and many others as a teenager.

 Artificial Intelligence is the theory and development of computer systems able to perform tasks normally requiring human intelligence, such as visual perception, speech recognition, decision-making, machine learning and translation between languages. At this point I’d like to highlight that we are not talking about thinking machines in the sense of machines with consciousness or self-awareness, merely machines that perform human-like tasks. There is also the important distinction between performing tasks and “understanding” those tasks. The other important definition is that of Machine Learning – often referred to as a subset of AI, machine learning uses statistics to develop self learning models.

In the AI field there are also two important distinctions – Narrow AI and General AI. Narrow AI refers to machines capable of completing specific narrow tasks – driving a car, analyzing an image, recognising speech, making a movie recommendation, estimating the probability of a legal case being successful and it is in this area that we are seeing so much progress. General AI refers to machines with the ability to apply intelligence to any problem, rather than just one specific problem. There is a fair amount of controversy as to how far away we are from General AI so for now we will focus on Narrow AI.

Both Oracle and SalesForce have made announcements this week of building Narrow AI functionality into customer applications. Oracle’s Adaptive Intelligence Applications are based on insights gleaned from Oracle’s Data Cloud, using anonymised web scale data and applying data science to target business outcomes, including next best offer, best-fit candidates and optimised payment terms.  The system is continuously learning and adapting as the end user interacts with the system. This highlights that the tipping point has been reached – products with embedded AI (such as Clinical Decision Support), AI as the basis for new products (like Siri or Quill) and AI-as-a-Service have all come of age.

So what are the implications for businesses? In a world where more and more products have embedded AI it can be a winner takes all! If your algorithms are better at attracting and retaining customers then you collect more data which improves the algorithms –the network effects are self re-enforcing! It’s not too hard to envisage a world where we have many brands of driverless cars on the roads but the AI or “brain” in each car comes from a single organisation, and that single organisation is collecting all the data from every driverless car to further improve the AI. It’s hard to be the number two competitor in a market where more data means better safety. To this point there are sure to be government policy initiatives to encourage or mandate the sharing of data from driverless vehicles (a topic for another day).

Embedding AI functionality into existing products provides differentiation that is difficult or near impossible for competitors to re-produce. Suddenly we move from feature/function differentiation to differentiation based on the quality of the data and the AI. Highly relevant in the software industry as this week’s announcements confirm.

So what should companies be thinking about? What are the opportunities for embedding AI into your products or processes? Do you want to be the consumer of AI from third parties or a creator of AI for your industry?  AI and big data are two clearly linked trends and for most of the customers I talk to it will be a natural transition from big data initiatives to AI initiatives, leveraging the investment in big data. The other interesting trend is AI- as-a-service, for those customers not interested in building a capability from the ground up. It is certainly time to be assessing the impact of AI on your business processes, business model and importantly on the products or services you produce. 

Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.
בואו נעבוד ביחד
צרו קשר