Approximately one in three restaurants will go out of business in its first year. For construction companies, that figure rises to 53%. But AI projects are the real heartbreakers: A Gartner study found that 85% are destined to fail “due to bias in data, algorithms or the teams responsible for managing them.” Unfortunately, the profound fear of missing out means many organizations are jumping into AI projects with both feet even though they don’t fully appreciate the scope of work involved. “The best way to ensure you are on the correct AI development path is to start your AI project without thinking about the models,” recommends Eran Shlomo, co-founder and CEO of Dataloop. “Most of the data that the AI needs to perform at its best ability is not available to the development team," he writes. “This creates a ‘chicken or egg’ problem: Businesses need production data to deliver a functional model, but the model needs to exist in order to go to production.” In a post aimed at non-technical managers and senior developers, he shares a framework for building a core team consisting of data scientists, domain experts and data engineers who can build a system that can learn from its mistakes iteratively. Via collaboration, “the AI provides automation, speed and low costs” while the team steers “the AI to a correct result in a constantly changing environment.” According to Shlomo, working along these lines generates a machine learning data flywheel, “essentially planning a learning system rather than an AI model that works properly at a single point in time.” Thanks very much for reading, Walter Thompson Editorial Manager, TechCrunch+ @yourprotagonist Read More |
Commentaires
Enregistrer un commentaire
Thank you to leave a comment on my site