As the clock ticked midnight in Seoul, Sumin, co-founder of Deeping Source Inc., was delighting in nachos and dipping sauce halfway across the world in the eastern US. This moment was special; it was the eve of launching our revolutionary enterprise.
We were on a quest: to christen our new venture with a name that embodied our mission—making every piece of data worldwide securely accessible without compromising privacy. We scoured the internet, tossing around numerous names, only to find that many were already claimed. Amid our brainstorming, the crunch of Sumin's nachos over the call broke my concentration. "What's on the menu today?", I curiously inquired. Her response, while seemingly ordinary, struck a chord, “Nachos with... Hey, how about Deeping Source?” It was a eureka moment. We were all about delving deep into data, the very 'source' that powers everything today. Our naming journey culminated, and a new adventure began.
Backtracking a bit, my maiden venture, Olaworks Inc., found its new home with the tech titan, Intel, in 2012. As we integrated our computer vision software into Intel's ecosystem, I was plunged into the exhilarating realm of designing cutting-edge AI accelerators. The shift from traditional machine-learning algorithms to data-driven deep learning was transformative, influencing every facet of the tech world. Suddenly, even modest systems were ingesting gargantuan data sets—be it articles, texts, or financial transactions.
For over half a century, AI has journeyed through innovations and evolutions. Yet, concerns about privacy, copyright, fairness, and ethics were mostly limited to science fiction—until recent AI breakthroughs. With the capability to process video feeds from countless cameras and pinpoint identities globally in real-time, modern AI thrust these concerns into the limelight. As people awakened to the goldmine that their personal data represented for big corporations, landmark regulations like GDPR and CCPA were set in motion, reshaping the AI landscape.
Being part of a multinational like Intel, I saw firsthand the effects of these regulations. Massive data sets, once the backbone of AI projects, were now obsolete. AI initiatives were rebooted from scratch. It became clear to me that the future of AI wasn't solely in hardware but also in the responsible management of data. This realization led me to leave Intel and sow the seeds of Deeping Source.
From day one, our team at Deeping Source embarked on a deep dive into data. We envisioned a world where data could be safely curated, accessed, and shared. Initially, our goal was a grand marketplace where individuals could seamlessly share their data with AI entities. But we pivoted to prioritize creating secure data sources and their analysis. Our innovations spanned a spectrum—from safeguarding privacy by extracting personally identifiable information, ensuring data copyright through watermarking, to pioneering privacy-centric AI development via anonymized data, which we proudly term 'Anonymized AI'.