In just a few months, Google will again head to Washington DC and meet with the government for a two-week remedy trial to cement the fate of what happens to Chrome and its search business in the face of ongoing antitrust court case(s).
Or, Google may proactively decide to make changes, putting the power in its hands to outline a suitable remedy. Regardless of the outcome, one thing is sure: there will be far more implications for AI than just a shift in Google's Search business.
A move of this magnitude will have far-reaching impacts across a broad spectrum of companies, enterprises, and consumers. It also has the potential to democratize access to user data and accelerate unaddressed AI development while addressing fundamental concerns about transparency and ethical AI usage.
Leveling the Data Aggregation Playing Field
As of now, Google Chrome equates to nearly 70% of the global browser market share. This provides Google and its parent company, Alphabet, unprecedented opportunities to access user behavior data at an unrivaled scale. For every search session, link clicks, and web browsing session, high-quality behavioral data is generated.
This data acts as the plumbing that impacts all of Alphabet's other existing products, such as Google Search and YouTube. It's been talked about for years, but these products exist within a walled garden, meaning competitors don't have access to these insights as the data is locked within a closed-loop system.
Making this data more accessible and transparent should be seen as a positive. Unlocking the existing walled garden will increase competition, lower barriers to entry, drive more innovation and enable better products for consumers.
As we look to the future, regardless of what remedies percolate, there will likely be more incentives for third parties to tap into this resource and build non-Google partnerships.
The Challenge with Bias, Transparency and Fairness
AI bias refers to human biases that enter AI algorithms via training data, leading to distorted outputs and potentially unfair or harmful outcomes. Biases are inherently present in all data because models inherit biases across the information they're trained on, much like humans. The first step in building a trustworthy AI model is to ensure you deeply understand the data.
But when models are created in a closed-loop system, you risk having products and models trained on limited data sets. You need to rely on the entity incorporating diverse viewpoints to reduce the risk of bias or hallucinations. This is a harder hill to climb and presents more significance for skewed recommendations and misinformation.
If or when access to Chrome becomes broadly available, companies can build models that reflect a wider range of user experiences, paving the way for more representative AI.
In this hypothetical scenario, models will become trained on more diverse, high-quality datasets that could reduce hallucinations and incorrect or misleading results generated by a model.
The best way to prevent hallucinations and biases in AI model building is to:
1. Ensure your data has not been compromised.
2. Use an AI platform with the validation tools to understand your model's behavior, contest the results, and optimize performance over time.
At its highest potential, potential remedies to breaking Chrome away from Google would give companies more control over their content's inclusion in AI models, preventing risks of hallucinations and bias.
What a Leveled Playing Field Looks Like
While Google isn't the only big tech player that operates as they do, its market share certainly puts it in a field of its own. Removing the competitive advantage and resetting the market share will go a long way for the future of the open web and create new opportunities to support growth in the tech enterprise ecosystem.
Having more significant amounts of data accessibility and breaking down a walled garden of this size has the potential to open more access to diverse data sets for SMBs, startups, and other independent enterprises working on bespoke AI applications.
Everyone sitting at the same table creates endless possibilities, encouraging smaller tech players to solve unique challenges in new ways and steer a different way forward.
How to Make Sense of It All
One thing is abundantly clear — change is coming. The DOJ's recommendation for Google to divest Chrome in any capacity will create a seismic shift across every facet of lives. That may be hyperbolic, but it doesn't feel any less accurate.
If, when, or how it happens will address ongoing monopolistic concerns. It will also provide a fundamental shift and opportunity to redefine how data powers innovation and its impact on our lives.
Any announced remedy will likely raise questions, uncertainty, and chaos. However, it will also provide an opportunity for all stakeholders — innovators, engineers, and others from companies of all sizes — to ensure that the tools they create incorporate clean and responsible data that can lead to more inclusive innovation and significant progress.