• AI isn’t failing, people are failing with AI AI isn’t the problem - rushing it into the wrong tasks without the right data, expertise or guardrails is what makes projects fall apart. • In 2018, Google released the AI modelBERT, forever changing how machines understood context in a language. • BERT, short for Bidirectional Encoder Representations from Transformers, solved a long-standing problem in natural language understanding. • Before BERT, researchers needed multiple bespoke models (and datasets) to understand the different contextual meanings of human languages. • BERT demonstrated that one model could process contextual meaning across multiple languages (via mBERT). • While BERT became a fundamental building block in natural language processing (NLP), its impact on how we interact with computers came from its application.

Article Summaries:

  • AI experts argue that the technology itself is not the problem-mis‑deployment is. The article cites Google’s 2018 BERT breakthrough, noting that its success came from domain‑specific expertise, close collaboration with researchers, and careful tuning rather than sheer compute. Today’s large language models (LLMs) such as GPT, trained on trillions of tokens, offer similar power but are “dangerously easy to deploy poorly.” McKinsey reports that many AI pilots stall or fail because decision‑makers lack a high‑level grasp of how the models work, the data they need, and the guardrails required for safe, scaled impact.

Sources: