By Tim Rollins
Few news topics—much less technology topics—receive as much attention as artificial intelligence today. Generative AI tools like Chat GPT, Google Bard, DALL-E, and MidJourney have been featured in headlines on the front page of the New York Times, the cover of Time Magazine, and on television programs like 60 Minutes. Predictions for AI’s possibilities range from utopian visions of the future to the potential for the end of humanity caused by AI run amok.
But AI has been around for a while now. The first generation of AI used rule-based approaches to perform tasks. Programmers hard-coded rules that computers could execute faster and more efficiently than humans. Rules-based AIs can be extremely effective at games, speech recognition, and textual analysis. Deep Blue, the first chess AI to beat a reigning world chess champion, was a rule-based AI.
The second generation of AI programmers embraced machine learning, a type of AI that uses algorithms to learn from labeled training examples or iterative cycles of prediction and analysis of outcomes. Machine learning has applications across many fields, ranging from spam filters and Netflix recommendations to predictive coding in the review phase of e-discovery. Modern machine learning algorithms are known as deep learning, meaning they take advantage immense computing power to assimilate vast amounts of data rapidly, to produce more “intelligent” and accurate outputs.
Generative AI, which is what most people think of when using the term artificial intelligence in 2023, is a subset of deep learning AI that can produce new content outputs based on their understanding of the patterns and structure of their input training data.
While generative AI has its uses in the realm of legal technology--for example generating copy for briefs or motions--we've heard far more about mishaps related to generative AI (such as here and here) than whatever time it might save for attorneys. Tried and true AI technologies, though, such as predictive coding and TAR (and TAR 2.0) all leverage machine learning and deep learning to deliver valuable time and cost savings--even when they are only used to support human document reviewers, rather than replace them. Even for technology skeptics, there are tactics organizations can use to derive benefits from AI without committing to it as the sole means of document review, including:
- Prioritizing documents for human review
- Privilege review
- Reviewing incoming productions
- Secondary review of lower-priority documents
- Quality control for human reviewers
Nonetheless, these older AI technologies aren't used as nearly as often or by as many as they could be. Whether you wholeheartedly embrace AI or are skeptical of it, the cost- and time-saving benefits it provides are increasingly difficult to ignore.
Check out our recent whitepaper E-Discovery Technology at a Glance: Artificial Intelligence to learn more and decide if you're ready to use AI--even if it is old-fashioned machine learning!