After months of hype regarding OpenAI and ChatGPT, Google has thrown its hat into the AI chatbot ring with Bard, an artificial intelligence search tool which may have profound implications for enterprises.
As outlined in a blog by Sundar Pichai, CEO of Google and Alphabet, Bard is an experimental conversational AI service powered by Google’s Language Model for Dialogue Applications. Known as LaMDA for short, the large language model has long been in the making at Google HQ, but without the public access that has seen ChatGPT become what one study has dubbed the fastest-growing consumer application in history.
Some impressed by ChatGPT’s uncanny skills in answering user prompts have nonetheless been awaiting to see what Google has up its sleeve; besides controversial claims from Googleplex insiders that LaMDA is an AI tool which has reached sentience(!), one can always assume that the search giant’s vaunted, veteran status in tech means it’s awash in data and forward-thinking technology which OpenAI can only dream of.
As ChatGPT arose in prominence towards the end of last year, it was surprising to see Google become a laggard in the AI stakes. Tools developed in secret remained behind the curtain, whilst Microsoft announced it would incorporate OpenAI’s tool into its Bing search engine. Redmond has been partnering and investing in OpenAI since 2019, making Google’s reticence look bad (or Bard) for business.
Yesterday’s Bard announcement may change all that – dwarfing even last week’s news of an almost $400m investment by Google Cloud in ChatGPT rival Anthropic – even if the tech is, for now, available to private testers in beta before a public rollout in the “coming weeks.” The secret sauce is becoming less secret as Google plans to onboard individual developers, creators and enterprises to test out its Generative Language API, initially powered by LaMDA with a range of models to follow thereafter.
Bard may also change things for enterprise tech. Firstly, it’s hard to miss that Google is forsaking company secrets which may chip away at its dominion in development. Alphabet is likely not, as some have commented, to be scared of releasing potentially world-changing AI without extensive testing. It’d rather keep its sauces secret, one would wager, especially with lacking financial results on its ledger.
The chatbot game will also change for organizations; internal-facing conversational agents will suddenly look very first-generation once the likes of ChatGPT and Bard run amok.
Theo Priestley, CEO and founder of AI analytics platform Metanomic, commented on LinkedIn that “SAP, Oracle, IBM, Salesforce, Infor, Netsuite, of course Microsoft and many others will all be scrambling to integrate either Bard (if Google allows it) or ChatGPT (as it’s already an open tool) to quickly offer enterprise-grade solutions that allow businesses to use the power of (large language models) pointed directly at their own data.”
Chatbots, Priestley reminds us, work when “fed and trained on specific and targeted information…(info) that’s usually wrapped up across large estates of enterprise software solutions like ERP or CRM.”
The CEO also asks which businesses can afford the volume API charges at scale and how they can make it work with current licensing revenues if they use either Google or Microsoft.
“Will it be more cost effective to develop and train their own LLM vs the immediate lead over the competition they’ll get paying the price elsewhere?” Priestly asks. “Time will tell but it’s going to get bloody.”
The battleground may indeed open up shortly, what with Microsoft hosting a surprise news event today in USA time predicted to announce major Bing/ChatGPT news.