May 23, 2023

AI: Three things to think about

Artificial Intelligence has morphed from science fiction to a robust, transformative tool. Deploying AI to revamp internal operations or to augment product features is becoming an effortless endeavor. An ever-increasing number of enterprises and start-ups are employing AI for product innovation and business problem-solving. It is estimated that over 60% of mid to large-scale companies worldwide already utilize AI.

However, it’s crucial to understand certain nuances before incorporating AI into your product. We’ll also discuss why uploading sensitive documents to a public neural network could be a perilous affair, and the intriguing role China plays in this evolving story.

Thing Number 1. The Reigning Titan of AI Research and Development

Innovation is the driving force behind progress in Artificial Intelligence. Since AI’s inception, the United States has held a leading position in this realm. Yet, it’s time to examine who currently holds the reins in this dynamic research and development landscape.

The name OpenAI, an American powerhouse, often dominates the conversation about AI. The organization has garnered extensive attention for its groundbreaking advancements in AI. However, in terms of research and development, the spotlight extends beyond the US to China. Despite a lower volume of venture capital investments, China leads the US in terms of AI-related scientific papers. Its academic output stands on par with the US in terms of relevance and citations. China boasts a larger pool of science and technology graduates and is gradually gaining ground in the quality of education, but lags in attracting international expertise.

Chinese policymakers envision AI as the primary catalyst for industrial modernization by 2025. By 2030, they aspire for China to emerge as the world’s AI innovation hub, investing a projected $150 billion in foundational industries and $1.5 trillion in associated sectors. Should current trends persist, China may soon wrest the AI research and development crown.

Thing Number 2. The Double-Edged Sword of AI Start-ups

Corporations increasingly hinge their business strategies on AI deployment, viewing it as a potential competitive edge for enhancing productivity and optimizing operations.

However, major tech firms do not monopolize the AI landscape. Emerging start-ups, leveraging open-source software, are rapidly gaining traction. Hugging Face’s multilingual language model BLOOM, and the buzz-worthy Stable Diffusion, an open-source AI model rivalling OpenAI’s DALL-E 2, stand as notable examples from 2022. The number of AI start-ups has multiplied 14-fold over the last two decades, with three hundred companies registered in Q1 2023 alone.

Despite larger corporations downsizing their teams due to gloomy global economic forecasts, generative AI start-ups are attracting venture capital interest. Indeed, investment in AI start-ups worldwide has surged six-fold since 2000, enabled by API access to neural networks. Start-ups are capitalizing on pre-existing neural network models and algorithms, focusing on user interface design and new product features rather than investing heavily in developing their own neural networks.

However, reliance on another’s technology poses risks: the plug could be pulled on API access, or contractual terms could change.

Yet another thrilling chapter in the start-up adventure involves idea-hunting expeditions from larger competitors. With its ecosystem of brands and a unique platform for product promotion, Amazon has been known to employ data to zero in on audience needs, spot market gaps, and refine their products accordingly. Once fine-tuned, these products are effectively “catapulted” onto their platform, nudging competitors towards the sidelines.

Similarly, a certain social network known to mimic popular features of competitor apps, erecting barriers to start-ups vying for a spot in the social networking arena. Quite the plot, isn’t it?

Hence, to avoid being swept up in this wave of competitive gamesmanship, it’s crucial to thoroughly assess all risks through analytics at the planning stage of market entry.

Thing Number 3. Navigating the Minefield of Corporate Security and Data Breaches

AI provides companies with the means to foster innovation, develop unique products, and bolster market competitiveness. Nevertheless, a dark cloud hangs over this positive aspect: potential data leaks. How do the perks of AI stack up against potential threats?

A hallmark of AI is its reliance on massive data volumes for learning. For instance, to generate an on-demand product image, the AI must first “digest” thousands of real images or drawings. But where does data privacy factor into this equation? Data uploaded to a neural network are made public, and therefore potentially accessible to competitors. Additionally, any data uploaded is no longer under your company’s intellectual property protection.

Envisage a scenario where your company uses a neural network to generate a document containing trade secrets, say a method for diagnosing genetic diseases. The training data would comprise real medical data, potentially falling into the wrong hands. Or suppose a public neural network generates a copy of a legal document, potentially allowing competitors or ill-wishers access to confidential information.

In response, companies are rushing to create regulations that restrict the use of neural networks and protect confidential data. Despite the uncertainty of these measures’ effectiveness, it seems inevitable that start-ups will emerge, promising secure corporate data handling.

In conclusion, as we navigate this brave new world, we must view these challenges not as deterrents, but as opportunities for growth and adaptation. The onus lies with companies to not merely restrict AI usage but to comprehend AI’s features and learn how to engage with it effectively and safely.

Share article