The Consumer Electronic Show (CES) was full of artificial intelligence (AI) agents of change this past week. Amazon noted that 28,000 products are now partnered with Alexa, up from 4,000 this time last year. Distributing more content is a key focus of AI home devices, and Amazon, Google, Microsoft, and Samsung were all showcasing the AI-enabled life-enhancing features of their digital assistants. For this trend to continue, we need to embrace the policy challenges that AI brings to data collection and privacy.
The explosion in AI products has been made possible by today’s network speeds. High speed networks enable applications to take advantage of real time information flow to deliver media, communications, and information such as live GPS data, then feed it back into cloud computing software to curate and manage the data more efficiently and accurately than ever before. Smartphone and voice assisted platforms are powering the app economy and these applications need the efficiency brought on by more data aggregation. AI has become an easily accessible technology for both large corporations and individual users, and society has become unknowingly dependent on AI and machine learning to make sense of the flood of information available.
AI is causing major changes throughout both small companies and entire industries thanks to the enhancements of data, connectivity, and computing resources merging with mobile, cloud computing, and platform technologies. In the past, these tech tools were expensive to deploy and only available to top industrial giants in sectors such as banking, energy, transportation, telecommunications, and healthcare. However, as this year’s CES showed, the benefits of AI are quickly filtering down into the consumer sector.
Cloud based companies are now offering AI capabilities as a service. Amazon, Google, IBM, and Microsoft are all working on advanced machine learning capabilities that can be built upon within the cloud network to allow analytical software to tailor products through automated learning processes. This will allow for more efficient product management, lowering the barrier to entry for new entrants into a market. Thanks to newer technologies available via cloud computing and better data with AI we will see smaller companies that can rent access, rent production, and rent consumer attention without the headcount and expenses of a major corporation.
AI’s ability to successfully tackle a problem is a function of the amount of data that it is exposed to: the more data it has seen, the more accurate its analysis can become. AI’s function is solving problems — it can make suggestions that allow humans to make better decisions with more efficiently organized data from known reference points as well as with real-time data fed into applications. It does this by seeing trends across large amounts of data, discerning patterns in behaviors and information flow that enable it to accurately predict behavior.
As Kai-Fu Lee points out in his recent book AI Super-Powers China, Silicon Valley, and the New World Order, “there is no data like more data.” This is an important point to take in as the US considers data protection legislation and looks at data protection components within national privacy laws. By limiting the collection of transaction data, for example, laws like the EU’s General Data Protection Regulation (GDPR) will make reviewing pertinent data to educate algorithms for AI very difficult. This could mean support for AI projects in Europe will decline with these government-imposed constraints.
Lee notes that “reliance on data for improvement creates a self-perpetuating cycle; better products lead to more users, those users lead to more data, and that data leads to even better products, and thus more users and data.” AI can bring this efficiency cycle to users in the same way that it has given whole industries the data to reimagine their processes, procedures, and expenditures.
There is room for a collaborative middle ground on AI productivity, data collection, and privacy. AI allows decision makers to use data to engage in higher value tasks that require human experience. As differential privacy tools that allow for disaggregation of personal information in data collection become more prevalent (for example, the technology that Apple uses to engage users for data without needing users’ actual identities), AI can take advantage of the economy of scale that enables machine training while still protecting individuals’ privacy.
As we continue to move forward into the age of data, we need to find the balance between individual privacy concerns and the ability to make information available for machine learning to reap the benefits of AI. The solutions demonstrated to date through the use of AI algorithms and the innovations brought about by mobile, cloud computing and technology platform tools has created a new level of information implementation that has benefited both industries and individuals.