A background in Arts isn’t the most conventional starting point for tech professionals. But that didn’t matter to Rodger Werkhoven when he made a career shift from the world of art and advertising to the frontiers of artificial intelligence (AI) development.
In this exclusive interview with Africa Hyperscalers, Werkhoven talks about his unique path, acknowledging the role of inspiring mentors and serendipity.
“I’m incredibly grateful for the fortunate twists of fate that shaped my path. Each twist, a seemingly random event, ultimately led me down this fulfilling creative journey,” Werkhoven says.
Three years ago his creative journey led him to join a small research lab, OpenAI, that was working on artificial intelligence and seeking creative minds with technological fluency. He became part of a global team of 30 individuals who helped train a model that could generate images from text.
Werkhoven has since moved on from that role, becoming a leading voice in the generative AI space who consults and advises companies such as Banijay on their GenAI efforts.
In this conversation, Werkhoven discusses AI’s role in revolutionizing Africa, exploring the broader implications of AI, particularly its transformative power for education and entrepreneurship in Africa.
This interview has been edited for brevity and clarity.
Tell us about yourself
I’m a blend of creativity and technology. You could call me a techie with an artistic side. I enjoy using my skills to make new things – music, art, and fresh concepts. What truly excites me is the fusion of these two forces. Many people lean heavily towards one or the other, but the combination of the two has always fascinated me. For 30 years, I’ve channeled this passion into creative director roles across various industries: software, advertising, and design. Over the past few years, I’ve explored the world of developing artificial intelligence models.
Your work experience shows that you’ve been in many creative roles, how did you pivot to technology?
My fascination with technology and creativity started when I was young. My dad’s Commodore 64, a home computer with rudimentary graphics and game capabilities, sparked my imagination. It could do animations, play games, and even make music – all in eight bits! This, coupled with my love for science fiction (books like Star Trek and Star Wars), fueled a dream of turning this passion into a career.
Luckily, an art academy in the Netherlands offered a unique program: full-time computer technology experimentation for aspiring artists. I jumped at the chance, to learn to program and develop a conceptual approach. Graduation opened doors to both art projects and collaborations. Software companies sought my help in designing user interfaces, while innovative advertising agencies saw potential in the then-nascent internet. They tasked me with creating internet concepts, alongside traditional commercial work like television and print ads. This combined experience, blending computer skills, creativity, and traditional artistic methods, shaped me into the versatile creative I am today.
Your journey reminds me of Steve Jobs, a brilliant tech leader who had a background in arts. How have you honed your skills over the years and how has this experience in the creative industry helped you?
Steve Jobs is a creative soul who captivated me and shared a similar journey. His biography revealed a childhood of tinkering with electronics and a fascination with aesthetics. He, like me, admired the minimalist designs of Dieter Rams, who brought sleekness to Braun’s everyday appliances. This resonated with Jobs, shaping his own design philosophy.
At art school, my teachers echoed these principles. They emphasized the importance of considering all design aspects, from functionality to user experience. “Keep it simple,” they’d say, “and prioritize the concept.” Form, they stressed, should follow function and idea. What message do you want to convey? What purpose does the creation serve? These lessons mirrored the approach I discovered in Steve Jobs’s work, solidifying my path. That chance indeed encounters with inspiring individuals can significantly impact your trajectory. In both our cases, influential teachers played a pivotal role.
Fortunately, when I was 18 years old, I wasn’t selected for the draft in the Netherlands (conscription for military service). This fortunate turn of events allowed me to pursue my artistic ambitions and attend art school. Perhaps the army would have steered me towards a career in mechanics or electronics, given my interest in technology. But fate intervened, and art school became a reality.
Ironically, the very program that fueled my passion was also the least popular at the school. Many students viewed computers as incompatible with artistic expression, leading them to other institutions. This niche program’s low enrollment was my gain. It offered a rare opportunity, as the school lacked experienced instructors in computer-based art. They relied on professionals from the commercial world, talented and entrepreneurial individuals who were themselves innovators. These exceptional teachers were another stroke of luck in my artistic journey.
I’m incredibly grateful for the fortunate twists of fate that shaped my path. Each twist, a seemingly random event, ultimately led me down this fulfilling creative journey. After graduating from art school, my first job was designing user interfaces for a software company. There, I befriended a young developer with dreams of Silicon Valley. We kept in touch, visiting each other across the Atlantic. As our careers progressed, we followed each other’s online endeavors.
Three years ago, my friend mentioned a small research lab called OpenAI, working on artificial intelligence and seeking creative minds with technological fluency. They were assembling a global team of 30 individuals to train a model that could generate images from text. Diversity was key – they wanted the model to reflect a range of cultural perspectives, not just American imagery. He offered to introduce me, and thanks to our long-standing friendship, I received an invitation from OpenAI to contribute my creative expertise as a European team member. This opportunity stemmed entirely from a chance encounter years ago, a testament to the power of connection.
Let’s talk about your time at Open AI
While I joined OpenAI in 2022, the organization itself has a fascinating history dating back to 2015. Founded by visionary minds like Ilya Sutskever, Sam Altman, and Elon Musk, OpenAI emerged from a concern about the trajectory of artificial intelligence. They saw existing prototypes as rudimentary but with immense potential for growth, particularly in the hands of a single dominant player like Google. This concentration of power, they feared, could lead to misuse.
Joining OpenAI exposed me to fascinating research. Their project, GPT (Generative Pre-trained Transformer), was a powerful language model, but their ambitions extended beyond. They envisioned Artificial General Intelligence (AGI) – machines with human-like intelligence – but achieving this required ensuring alignment with human values. Unaligned AI could perceive us as competitors, a scenario we desperately wanted to avoid.
That’s where “DALL-E,” the project I joined, came in. DALL-E wasn’t intended as a creative tool initially. Instead, it served as a test for GPT’s alignment. Humans excel at visual processing; we can readily detect inconsistencies in images. Text, a later invention, is less intuitive. Therefore, DALL-E used image generation to assess GPT’s understanding. If you requested a “crocodile on a bicycle” and received a “koala bear on a comfy couch,” it would signal a misalignment in GPT’s interpretation. DALL-E also helped identify potential biases within the model. Since these models were trained on vast amounts of internet data, which can be rife with biases, DALL-E provided a valuable tool for mitigating such issues.
Our mission at OpenAI was to ensure AI benefited humanity. This meant safeguarding against biases and harmful content. DALL-E played a crucial role in this process. By analyzing the model’s image generation, we could identify and remove biases like gender inequality or religious intolerance. DALL-E essentially functioned as a quality control tool, ensuring the AI aligned with our values. The initial goal was to dismantle DALL-E once we were confident in the model’s alignment.
However, Microsoft’s investment in OpenAI opened doors for commercial applications. While competitors like Midjourney were already making strides with their closed-access models, OpenAI saw the potential for responsible commercialization. The lessons learned from DALL-E’s development proved invaluable in launching ChatGPT, which became a resounding success. In essence, DALL-E served as a blueprint for introducing powerful AI tools to the world in a responsible and impactful way.
How was your experience working with the Open AI team?
It was fun. The small, close-knit team at OpenAI fostered a truly collaborative environment. Open communication flowed freely, allowing us to share discoveries, biases we encountered, and user interface suggestions. Ideas were rapidly prototyped by the developers, creating a sense of shared creation. Natalie, the team leader, played a pivotal role. Warm and nurturing like a mother, she coordinated team onboarding and fostered a culture of acceptance. This wasn’t about rigid structures or conformity; everyone, including myself (an ADHD individual), the possibly autistic developers, and the likely ADHD but wonderfully chaotic Natalie, was valued for their unique contributions. Diversity was embraced. Skin tones ranged from mine, which I sometimes wish had more color, to those far deeper. We had Asians, women, men, and even those transitioning genders. OpenAI’s mantra – “talent, regardless of source” – was evident in our composition. During visits to places like Dubai, I’d be asked why OpenAI attracted such brilliant minds. My response: Unlike some countries, we value talent, even if it comes from homosexual individuals who might face societal resistance elsewhere.
OpenAI’s strength resides in its team. The small size fostered a collaborative environment where open communication was key. We shared discoveries, biases, and UI suggestions, and developers rapidly prototyped our ideas.
What are the opportunities and concerns that you have for AI?
AI presents the most human-like technology ever created. It mirrors our communication style, allowing interaction through text chat or even voice commands on mobile devices. This accessibility extends to languages like Nigerian dialects, reflecting its vast internet-derived knowledge base. This ability to mimic human communication holds immense potential, particularly in education. Imagine a future where these large language models, currently housed in massive servers, are miniaturized. Several companies are already demonstrating ultra-compact models suitable for mobile phone integration. Samsung’s recent launch and Apple’s rumored iPhone with built-in language models are testaments to this trend. The cost will also plummet, making education accessible to those who previously lacked the resources or literacy skills. This technology empowers individuals to develop entrepreneurial ideas and improve their communities, potentially positioning Africa as a leading continent due to its abundant human potential.
However, the potential for misuse is undeniable. AI’s vast teaching capabilities are not without risk. China exemplifies this concern with its AI-powered facial recognition system. Crossing a red light can trigger a decrease in your social credit score, potentially impacting your children’s educational opportunities. This misuse of AI technology is a stark reminder of the importance of responsible development. The potential for misuse lurks alongside AI’s immense potential. Consider the troubling possibility of weaponized drones equipped with facial recognition. These commercially available drones, originally intended for recreation, could be repurposed for targeted attacks. This dark scenario underscores the importance of responsible development.
History offers cautionary tales. Nuclear technology once hailed for its ability to generate vast energy resources, also yielded atomic bombs, nearly leading to global catastrophe. Like any powerful tool, AI requires careful handling.
To mitigate these risks, some companies are taking proactive measures. OpenAI, Anthropic, Microsoft, and others established the Partnership on AI (PAI), a self-regulatory organization dedicated to ensuring the ethical development and deployment of AI.
The approach to AI regulation varies. The US, championing free markets and innovation, hesitates to impose strict regulations. Europe, on the other hand, adopts a more cautious approach, implementing stricter rules to protect citizens’ privacy. For example, facial recognition technology is generally prohibited for commercial use in Europe, with exceptions only for law enforcement in specific situations. This highlights the diverse approaches to governing AI development.
How do you think AI can have the most significant impact on African infrastructure?
Education, given the abundant human resources on the continent. By educating the new generation, we can foster the emergence of professionals across various sectors, including engineering and data management. Nigeria exemplifies this shift, as it lacked legacy technology and swiftly embraced affordable innovations like mobile technology. This ingenuity led to groundbreaking advancements in banking, financing, and mobile banking solutions. Similarly, education must evolve. Traditional methods are no longer feasible due to resource constraints. However, with mobile devices and embedded AI, internet access becomes unnecessary. Companies like Rabbit are already paving the way with cost-effective devices. These devices, equipped with AI teachers, can revolutionize education without relying on internet connectivity.
By subsidizing these devices, banking and telecom companies can catalyze an educational revolution in Nigeria. This is feasible now because there’s no reliance on legacy technology or large data warehouses. This shift in approach is key to unlocking Africa’s educational potential.
We must also recognize that building numerous data warehouses everywhere isn’t advisable due to the significant cooling requirements and the scarcity of drinking water. Instead, the focus should shift towards deploying educational devices. Once a substantial number of individuals are educated and skilled as engineers, efforts should concentrate on developing infrastructure to connect them. This doesn’t necessarily entail traditional methods like laying cables along highways; rather, the decreasing cost of satellite technology presents a viable alternative. Therefore, while there’s immense potential for infrastructure development, it must begin with a reevaluation of education. Investing in large schools with hundreds of students per teacher is outdated and inefficient. Emphasizing personalized teaching methods would be a more cost-effective approach to educating Africa.
Aside from these, do you see any other opportunities?
In every domain where entrepreneurship flourishes, artificial intelligence offers vast potential. Whether in medicine, where AI can expedite drug discovery by identifying beneficial protein and molecule groups, or in the health sector, where entrepreneurial individuals outside of universities can leverage AI for innovative solutions, its impact is significant. AI streamlines processes, eliminating the need for decades-long research cycles to develop new treatments. Similarly, in the creative industry, AI empowers individuals to establish advertising agencies single-handedly, performing tasks that would typically require a team. This democratization of creativity challenges established companies to adapt and encourages new ventures to emerge. Moreover, AI facilitates innovation across various fields by enhancing efficiency, accelerating procedures, and fostering the development of eco-friendly materials derived from natural resources. Thus, regardless of the entrepreneurial pursuit, artificial intelligence, with its vast repository of human knowledge, stands as a powerful ally.
How do you think AI can be implemented responsibly in Africa’s digital infrastructure targeted to ensure job security while fostering inclusive growth?
What struck me about the Nigerian people is their strong sense of community and mutual support. During my time in Nigeria, I observed how the middle class actively engaged with and supported their less privileged neighbors, building beautiful homes while still living alongside them. This symbiotic relationship, where the middle class purchases goods from those not yet in their ranks, exemplifies their collective progress. This community-centric ethos is vital even in the integration of artificial intelligence. Take, for example, OpenAI’s DALL-E 3, capable of creating photorealistic images.
However, it’s essential to consider the impact on photographers’ livelihoods. While some AI developers may prioritize immediate capabilities, others may uphold ethical considerations. Just as societal norms dictate not stealing or harming others, we must establish ethical guidelines in AI usage. Government intervention is crucial, but individual responsibility and entrepreneurship are equally important, especially as certain jobs become obsolete or undesirable.
For instance, the role of call center agents handling product complaints may diminish because working in a call center and dealing with a barrage of complaints all day can be exhausting and demoralizing. Instead, utilizing ChatGPT to handle such mundane tasks could vastly improve the work environment for call center employees. This shift would allow them to focus on more meaningful interactions, providing quality customer service to brands that prioritize human connection. Companies that recognize the value of genuine human interaction may even be willing to pay higher prices for their products or services, thus preserving jobs that require a personal touch. Just as industries like asbestos factories have become obsolete due to health concerns, jobs impacted by AI automation will require adaptation and reinvention.
History has shown that technological advancements often lead to the evolution of professions rather than their extinction. For example, the invention of the steam engine or the camera initially threatened certain occupations but ultimately spurred innovation and creativity in those fields. Similarly, painters adapted their styles in response to the invention of the camera, leading to the emergence of new artistic movements. Therefore, in the face of AI automation, creativity and adaptability will be essential for individuals seeking to thrive in evolving industries.
What challenges and opportunities do you see for AI, in transforming Africa’s digital landscape?
We discussed the immense potential of using AI to address the challenge of educating people, especially in regions where building enough schools and securing an adequate number of teachers is impractical. However, there are also significant challenges to navigate. One such challenge is the risk of individuals exploiting AI technology for personal gain at the expense of their communities. If individuals become overly eager and prioritize personal profit over community benefit, it can lead to harmful outcomes at an accelerated pace. Therefore, fostering a culture of accountability, discussion, and collaboration is crucial to ensure that AI advancements serve the collective good.
Looking ahead, the next wave of innovation lies in Language Processing Units (LPUs), akin to the Graphical Processing Units (GPUs) and Central Processing Units (CPUs) we have today. This presents an opportunity to bring manufacturing capabilities to Africa, particularly to countries like Nigeria, which are rich in intellect and entrepreneurship. By establishing manufacturing plants in Africa, we can reduce dependence on external sources like China and empower African stakeholders to drive economic growth. However, it’s essential to approach this opportunity with caution to prevent exploitation or the replication of past mistakes, such as the Foxconn controversy in China.
In all, education remains paramount in this endeavor, not only to equip individuals with the skills needed to harness AI but also to ensure they understand the ethical considerations and potential pitfalls. By leveraging AI for education and simultaneously investing in local manufacturing capabilities, Africa can take significant strides toward economic self-sufficiency and prosperity.