Monday, December 23, 2024

Sundar Pichai wants real world results for his AI bots—so Google gave bonuses and golden bomber jackets to staff who came up with winning prompts

Must read

Sundar Pichai and his team have gone back to basics in order to rapidly improve the effectiveness of their AI products in the real world—by launching an internal prompts competition and handing out prizes to the winners.

Alphabet-owned Google reportedly launched a ‘golden prompt’ competition for its Gemini workspace product, which is home to a host of AI-enabled productivity and collaboration tools such as emails, calendar, meet and more.

‘Golden prompts’ are the instructions which can be fed to bots in order to achieve an efficient and accurate outcome.

Those who came up with the winning prompts were handed prizes such as golden bomber jackets with a white Gemini symbol stitched onto the back, or spot bonuses, per Business Insider citing internal documents.

Google did not immediately respond to Fortune’s request for comment, and did not confirm the volume or value of the prizes distributed.

According to BI, Google framed the competition to software employees with an example of a small business owner attempting to sift through all their online reviews left on Google maps.

The winning prompt in this scenario, BI reported, was: “Every time I get a new Google review from a customer, add it to my Reviews Sheet and suggest a response. Let me review the response before publishing to Google maps.”

Seven winners of the competition were picked, with their prompts spanning examples from tracking and following up business leads through to flagging customers who need an urgent response. Other cases included keeping track of invoices and receipts, as well as helping users find research on leads.

Defining exactly how Google’s AI products will work for consumers is front and center on the agenda for Pichai. In the company’s earnings call Tuesday, Pichai told analysts: “On the cloud and AI stuff, obviously, it’s something which I think will end up being a big driver over time. 

“You are definitely seeing early use cases. But I think we are in this phase where we have to deeply work and make sure on these use cases, on these workflows, we are driving deeper progress and unlocking value, which I’m very bullish will happen. But these things take time.”

Prompting 101

The latest slew of use cases follows an April update from Google which coached customers on how to use AI-powered Gemini in their workspace.

Given the speed at which AI has been adopted since a certain OpenAI product launched onto the market, the document may be necessary.

The public has being playing catchup with Big Tech’s AI race, coming to terms with the likes of ChatGPT, Bing and Gemini.

A study published by Pew Research last year found the average American could generally identify the use of AI—but mainly if it related to wearable tech which tracked health and exercise.

However 69% of respondents didn’t realize AI was in a slew of other products—from email services to music recommendations—suggesting a gap in understanding.

As a result Google has some pretty simple instructions for users. It writes: “Prompting is an art. You will likely need to try a few different approaches for your prompt if you don’t get your desired outcome the first time.

“Based on what we have learned during our Workspace Labs program, the most successful prompts average around 21 words, yet prompts people often try without knowing this are short—usually less than nine words.”

Some aspects of the Gemini rollout haven’t come off without a hitch, and the ‘Prompting 101 Guide’ makes it clear users may still get unexpected results. It adds: “Before putting an output from Gemini for workspace into action, review it to ensure clarity, relevance, and accuracy. And of course the most important thing to keep in mind: Generative AI is meant to help humans but the final output is yours.”

Impressing analysts

Despite the emerging technology’s penchant for surprising its creators, the market remains impressed by Google’s performance.

In its Q2 results Alphabet reported revenues of near-$85 billion, driven by search and cloud services, up 14% year on year.

The results led Goldman Sachs to maintain their ‘buy’ rating on Alphabet following the update. In a note published yesterday, analysts Eric Sheridan and Alex Vegliante wrote that continued revenue momentum and “positive commentary” relating to Gen-AI products were an upside for the stock.

They added: “We see Alphabet as the leading collection of AI/machine learning-driven businesses that is uniquely positioned to capitalize on the blurring of the lines between advertising, commerce and media consumption in the years ahead and rising utility across a number of computing platforms including consumer desktop, consumer mobile & enterprise cloud computing.”

Recommended Newsletter: The Fortune Next to Lead newsletter is a must-read for the next generation of C-suite leaders. Every Monday, the newsletter provides the strategies, resources, and expert insight needed to claim the most coveted positions in business. Subscribe now.

Latest article