AI21 Labs has recently introduced a novel question-answering engine called “Contextual Answers,” designed to enhance the performance of large language models (LLMs).

This new engine enables users to upload their own data libraries, which effectively constrains the LLM’s responses to specific information, thereby increasing trustworthiness and usability.

The introduction of AI products like ChatGPT has transformed the AI industry. However, many businesses remain hesitant to adopt such technologies due to concerns about their reliability.

Research indicates that employees spend a significant portion of their workdays searching for information, making chatbots with search capabilities a valuable proposition.

Unfortunately, most chatbots lack the sophistication required for enterprise-level applications.

AI21 has addressed this issue by creating Contextual Answers, which bridges the gap between general-use chatbots and enterprise-level question-answering services.

Users can now incorporate their own data and document libraries, enabling more specialized and accurate responses without the need for model retraining.

READ MORE: 2023 Ranking: 4 Best Crypto Projects To Invest In

This approach significantly reduces the obstacles to AI adoption that many businesses face, including high costs, complexity, and mismatches with organizational data.

One of the major challenges in developing effective LLMs, like OpenAI’s ChatGPT and Google’s Bard, is training them to express uncertainty when they lack sufficient information to provide factual answers.

Instead of admitting they don’t know, LLMs may “hallucinate,” generating fabricated information that doesn’t exist in their datasets, much like humans seeing things that aren’t there.

AI21 claims that Contextual Answers eliminates the hallucination problem by either providing relevant information based on user-provided documentation or refraining from giving any response at all.

This ensures that the AI output remains accurate and avoids misleading users with erroneous information.

Sectors like finance and law, where accuracy is paramount, have had mixed results with generative pretrained transformer (GPT) systems.

In finance, experts remain cautious due to the potential for hallucinations and information conflation, even when GPT systems can access the internet and external sources.

In the legal sector, a lawyer was recently sanctioned for relying on outputs from ChatGPT during a case.

AI21’s data-frontloading approach and intervention to prevent hallucinations offer promising solutions for these sectors.

The financial industry, especially fintech, may see increased adoption of GPT technology, which traditional institutions have been hesitant to embrace.

Similarly, the cryptocurrency and blockchain communities, which have had limited success with chatbots, could benefit from AI21’s novel approach.

Overall, AI21’s Contextual Answers represents a significant step towards improving the reliability and usability of LLMs, opening up new possibilities for their adoption in various industries where accuracy and precision are crucial.

Other Stories:

Bitcoin’s Tight Bollinger Bands Point to Impending Volatility: Analysts Predict Significant Price Movement

Vermont Department of Financial Regulation Issues Stark Crypto Warning

China’s Digital Yuan Soars: Business Travelers Can Now Pay for Flights with CBDC