Opening the Black Box of AI with Jonathan Wright

Source De[Code]’s deep dive into artificial intelligence culminates in the podcast’s third episode when host Ben Coffin sits down with renowned AI expert and chief technology evangelist at Keysight, Jonathan Wright. Wright’s authority in this space and keen ability to make complex technology digestible and approachable has made him a sought-after speaker for industry presentations, AI consortiums, podcasts, and TEDx talks. More importantly, Jonathan is involved in the standards-setting bodies establishing guiding principles governing the development and use of this important emerging technology. Source De[Code] listeners who are not yet familiar with Wright will easily understand why he is such a sought-after voice in AI after listening to his thoughtful, entertaining conversation with Ben.

Right out of the gate, we learn that Wright lives in Oxford, England close to Bletchley Square where Alan Turing turned the world on its ear by introducing the imitation game by proposing the simple question, “can machines think?” It is also apropos that he has made his home there and dedicates so much time to deconstructing the myths around AI that have grown out of the countless popular culture narratives that use the Turing Test as their primary plot device.

AI in Software: A Paradox

In the episode, Ben and Jonathan talk at length about how this concept of intelligent machines has resulted in marketing misrepresentation diluting the true power of artificial intelligence—especially in the software space.

Jonathan points to a recent Forrester report commissioned by Keysight which identified a paradoxical relationship between the interest in and adoption of AI in the software space. It suggests there is a willingness to embrace AI-enablement in furthering the digital transformation, but few have begun to do so. The report indicates that 45% of software executives are considering adopting AI for their software within the next three years. This contrasts with the 11% of technology firms that are currently using it today.

Jonathan observes that we are “in an age of people slapping AI on a product” to describe its overall functionality when “it is only one subset of a small function that may use an AI algorithm that has been developed”. The hype surrounding AI and the inevitable use of it in marketing to sell AI-enabled products as a magic bullet solution to every imaginable challenge. Understandably, software executives are approaching this promise with skepticism, resulting in the curious paradox.

Demystifying AI

At the root of this paradox is a misunderstanding about what AI is, how it works, and where the technology is in its development. AI technology is still in its infancy. While individual AI algorithms are excellent at sifting through data within a narrowly scoped ask, these algorithms are not as well suited to more general requests. This will not continue to be the case for long, though. The advent of programs like Dall-E and Chat GPT has prompted an increasing number of conversations looking at the existential threat that AI poses; conversations that are critical to have now while AI technology is still in its early stages.

Because artificial intelligence is perceived as a black box, it is easy for fear to creep into conversations and perceptions of the technology. The ethics of this technology is being discussed in many places. Educators are questioning how to guard against students using AI to assist in writing papers. Agencies like the Pentagon and AI consortiums are focused on establishing ethical guidelines that must be adhered to in the development of AI even as corporations like Google are eliminating their “anti-evil” teams which are tasked with ensuring the technology they are developing benefits society and isn’t simply being developed to test the boundaries of what is possible.

In establishing these guidelines, it can be hoped that we will avoid a Skynet situation. Equally as important, however, is opening the black box of emerging technologies and deconstructing the hyperbolic myths that arise in the absence of easy to comprehend information about their development and use cases. Keep tuning into Source De[Code] to stay in the know and help inform your networks about the reality governing buzz worthy technology like AI.

Meet Jonathan Wright, Chief Technology Evangelist at Keysight Technologies

Keysight's Chief Technology Evangelist and reknowned AI expert Jonathan Wright is the guest in episode four of Source De[Code].

Jonathon Wright is a strategic thought leader and distinguished technology evangelist. He specializes in emerging technologies, innovation, and automation, and has more than 25 years of international commercial experience within global organizations. He is the Chief Technology Evangelist and heads up Product Engineering (R&D) for Eggplant a Keysight Technologies company.

Jonathon combines his extensive practical experience and leadership with insights into real-world adoption of Cognitive Engineering (Enterprise A.I. and AIOps). Thus, he is frequently in demand as a speaker at international conferences such as TEDx, Gartner, Oracle, AI Summit, ITWeb, EuroSTAR, STAREast, STARWest, UKSTAR, Guild Conferences, Swiss Testing Days, Unicom, DevOps Summit, TestExpo, and Vivit Community.

Throughout the pandemic, Jonathan has volunteered as the QA advisory lead at MIT for the COVID Paths Check Foundation. He also sits on several notable committees, including the Harvard Business Council, A.I. Alliance for the European Commission, chair of the review committee for the ISO-IEC 29119 part 8 “Model-Based Testing”, and part 11 for the “Testing of A.I. based systems” for the British Computer Society (BCS SIGiST). Jonathon also hosts the Canada-based QA Lead and is the author of several award-winning books. His latest wordk is co-written with Rex Black on ‘AI for Testing’.

Profile Questions:

What was the 'aha' moment that started you down the path/influenced your journey to where you are right now?

During an internship, my first ever role in in 1990s was manually testing a PABX telco platform, I was provided with over 100 pages of printed out test cases (i.e., Phone A rings Phone B gets put on Hold then transfers Phone A to Phone C) after spending a grueling 6 weeks, I turned up to the boss and explained that I had tested every single test and they all passed … honestly… I did not just simply ticked the pass boxes. The reply was simply, ‘I believe you, every test has passed, every release for the last 6 years’ I then asked ‘why do you test it then?’ which they responded because we need too!

I was then lucky enough to become hired as a full time automation engineer was for Siemens R&D, I was lucky enough to find a mentor (Dave Deboskey) who had spent the last decade building and designing automation frameworks and had a number of patents on keyword, data-driven frameworks. After an intensive few week’s training in Boca Raton he left me with a single word of wisdom …. Gottawannawin (GOT-TO-WANNA-WIN) …. in the early days of test automation, if you got stuck, you had to work it out for yourself or direct peers within the company as the internet had very little around community (with the exception of Mercury) and you had to relay on a handful of experts.

If you hadn't chosen your current profession, what would you have pursued instead? Why?

I wanted to become an actor, but I’m not sure I could have coped with the rejection, I did get to do some cameo work in a Marvel movie so thankfully I can tick that box too!

Where can we find you when you're not innovating the future of technology?

In the metaverse, or a digital twin of myself!

Listen to Source De[Code] Today

Up next on Source De[Code], Ben tackles digital twins with a series of experts. For more information about the podcast, additional resources, and a chance to win Source De[Code] swag, check out the webpage.