The AI boom is getting more specialized

You know what they say: bot of all trades, master of none.

What happened: Former Salesforce co-CEO Bret Taylor and ex-Google VP Clay Bavor brought their AI startup Sierra out of stealth mode. Its conversational AI is geared towards customer service, but unlike chatbots that just pull lines from an FAQ, it can also take actions, like upgrading a subscription or managing a delivery.

Why it matters: It’ll be tough for new companies to catch up with OpenAI, Nvidia, or Big Tech companies in generalized AI. But Sierra is one in a class of startups showing there are opportunities for companies deploying their technology in specialized cases.

Zoom out: A growing number of specialized AI startups have caught investors’ attention just in the last few weeks.

• Magic AI secured US$117 million for its AI coding and engineering assistant.
   
• Qloo, which uses AI to recommend entertainment and predict cultural trends, raised US$25 million.
   
• Alembic wrapped a US$14 million Series A round for an AI platform that better measures return on investment for marketers.
   
• Ambience Healthcare and othercompanies that use AI to handle medical records are in high demand among VCs.

In Canada: One potential competitor with OpenAI and Anthropic in more generalized AI is Toronto-based Cohere. But its non-profit research arm is also releasing models like Aya, a multilingual LLM meant to address a language gap coming from AIs being mostly trained on English and Chinese.

• Toronto’s Untether AI is building more energy-efficient chips. It recently named an Intel veteran as its new CEO, and added a former Google and Nvidia engineering executive to its board.
   
• Montréal-based Sibli operates an AI geared towards managing investments. It secured a US$4.5 million seed round this month.

Big picture: The big players are exploring more specialized AI applications, too. Nvidia plans to set up a unit to build custom chips for specific AI tasks, in addition to its generalized chips. OpenAI is reportedly branching out into agents that will control your apps and devices for you, as well as a search engine. Google and Apple are developing AI coding assistants.