Alphabet results reveal extent of AI-powered programming
The parent company of Google is using its Gemini generative artificial intelligence tool to code a quarter of new software development initiatives
During an earnings call for the third quarter of 2024, Alphabet CEO Sundar Pichai said the company would make its Gemini artificial intelligence (AI) model available to developers.
The company reported a 15% increase in revenue for the quarter, reaching $88.3bn. Google Cloud posted a 35% revenue increase to $11.4bn, led by accelerated growth in Google Cloud Platform (GCP) across AI infrastructure, generative AI (GenAI) and core GCP products.
“The momentum across the company is extraordinary. Our commitment to innovation, as well as our long-term focus and investment in AI, are paying off, with consumers and partners benefiting from our AI tools,” said Pichai.
During the earnings call, Pichai discussed how AI was helping Alphabet streamline software development. “We are organising the company to operate with speed and agility. We’re also using AI internally to improve our coding processes,” he said.
According to Pichai, this is helping Alphabet to boost productivity and efficiency. “Today, more than a quarter of all new code at Google is generated by AI, then reviewed and accepted by engineers. This helps our engineers do more and move faster,” he said.
Pichai said the Gemini team has now moved to Google DeepMind, to speed up the deployment of new models and streamline post-training work. This, he said, follows other structural changes that have unified teams in research, machine learning infrastructure and development.
In her prepared remarks, incoming chief financial officer Anat Ashkenazi said: “As I look at the business, I see opportunities for further growth, propelled by AI. I believe that we are well-positioned to deliver meaningful innovation, which will translate to revenue, given our strength in the core pillars that are required to succeed in AI at scale.”
When asked about the company’s AI infrastructure investment, Pichai said: “We have all the leading AI accelerators – GPUs [graphics processing units], TPUs [tensor processing units], as well as CPUs [central processing units], and we are investing in all of them.” He added that the company was also receiving Nvidia’s latest hardware.
Looking at the capital expenditure on AI infrastructure, Ashkenazi said the majority of this spending has gone into technical infrastructure. “Approximately 60% of that investment in technical infrastructure went towards servers, and about 40% towards datacentres and networking.”
One of the areas Ashkenazi said she wants to address is improving margins in the company’s GCP public cloud business. Answering a question on making the business more competitive against rival cloud providers with better margins, she said: “As we scale the business, we have more opportunity for margin expansion, but what shouldn’t be underestimated is the work the team has done to drive efficiencies across the cloud business.”
This, she said, has involved headcount management, facilities management and other process efficiencies that “go to the bottom line”. Yet while she praised the work to re-engineer the cost base, she said “any organisation can always push a little further”. One area she is looking to push is using AI within Alphabet’s own business processes.
Commenting on the results, Forrester principal analyst Tracy Woo said: “Cloud and AI continue to be big growth drivers for Alphabet as YoY [year over year] targets are exceeded by over a quarter point. As the company continues to remake itself as the AI leader through massive investment in cloud AI offerings, and to enhance its search capabilities, expect stronger traction with cloud in the coming year.”
Read more about AI for coding
- Amazon Q Developer tool has helped BT write 200,000 lines of code.
- At GitHub and for many coders, AI technology can shoulder some of the cognitive burden. However, inaccuracies that come with the tools can be hard for some.
Originally published at ECT News