Google or Microsoft?

While?NVIDIA ?clearly is an undeniable frontrunner in the AI race, the competition between?Google and Microsoft? seems to be only intensifying – the verdict on who is truly better at AI remains to be seen. Will Microsoft’s commercial acumen trump Google’s model superiority?

A few days ago,?Vin Vashishta , founder of V Squared shared an?interesting thought ?where he said amid the intense competition between the two, the success of AI products hinges not just on technical capabilities, but also on effective product management.

He said that data and AI product managers are emerging as crucial players in this race, ensuring that monetisation, production, and commercialisation are prioritised from the get-go, instead of being an afterthought. “A competitor that is better at delivering solutions to customers will win even with less advanced capabilities and models,” he added, saying that?Google is better at AI ?than?Microsoft . But, the latter is strategising it better.

The approach to decision-making is quite unique for both.?Dipanjan Banik , a principal product manager at?Microsoft ?said that Google’s decisions tend to be more engineering-driven, with a focus on developing new technologies and capabilities that can be integrated into its products and services. “In contrast, Microsoft’s decisions are more product-driven, with a focus on delivering innovative products and services that meet the needs of its customers,” he added.?

That doesn’t mean that Google isn’t top of its league. The company?claims to be mindful ?of releasing its technology into the market. Previously, it has made significant contributions, including?Vertex AI ,?Google Maps ,?PaLM , etc.?

Amid the ongoing battle between Microsoft and Google,?NVIDIA ?is emerging as a clear winner — i.e. because nearly all the existing SOTA models are trained on NVIDIA GPUs, alongside providing infra support for cloud services.?

However, the ambitious Microsoft is looking to outplay the chip giant in its own game. The company is reportedly working on its own AI chips called ‘Athena’ that can be used to train LLMs and avoid costly reliance on NVIDIA, where it could possibly slash by a third.?Read:?Microsoft Joins the Chip-munks, Will Make its Own AI Chips.

Top Stories of the Week >>

AI Startups Need to Learn Stability Tricks from Databricks?

The?Forbes annual AI 50 list ?missed to notably include Emad Mostaque’s Stability AI, whose open-source image generator Stable Diffusion tool became wildly popular after its release last year, while Databricks, which was decidedly working in an ‘unsexy’ segment in tech, secured a spot.

The decision surprised many as was evident on Twitter, but served to show that glaring difference between the two.?Read more here.?

Why is ETL Dying??

Organisations are moving away from ETL thanks to the emergence of AI agents such as?AutoGPT and AgentGPT , alongside the preference for unstructured data lakes over data warehouses. Traditional ETL pipelines have transformed to meet the demands of modern analytics use cases, but they can still be viewed as a hindrance to better performance and require substantial resources. AI agents, on the other hand, used properly can automate data processing and analysis, providing faster and more accurate results.?Read more here.?

Replit Knows India Better than Amazon, Microsoft

Last month, Google?partnered ?with Replit with the aim of competing with the likes of Microsoft’s GitHub, Amazon CodeWhisperer and OpenAI. In our recent interaction with the India head at Replit,?Anushul Bhinde ,?AIM?noted that the company is way ahead of its competitors when it comes to providing accessibility to budding software developers in India.?Read the complete story here.

AIM Videos >>

The Curious Case of Raja Koduri's Departure from Intel?

In our latest episode of ‘Story Kya Hai,’ AIM brings you an exclusive interview with former Intel AI chief architect,?Raja Koduri , discussing his switch to starting a new venture and more.?

AIM Events >>

Workshop: Accelerating Deep Learning Inference Workloads at Scale

NVIDIA, alongside AIM, is organising a free webinar on April 25, 2023, between 3 and 4 PM IST. This webinar would be introducing folks to?accelerate deep learning inference workload at scale –a must-know set of skills for developers in the generative AI boom.?

Register Now!

AIM Shots >>

Cloud computing provider CoreWeave recently raised?$221 million in Series B funding ?led by alternative asset manager Magnetar Capital, with contributions from NVIDIA, Nat Friedman, and Daniel Gross.?

AI-powered Analytics company?ThoughtSpot ?has acqui-hired Bangalore-based analytics and consulting services company?SagasIT Analytics .?Read more here.?

After Tim Cook’s recent visit to India, inaugurating two official Apple stores in Mumbai and Delhi, alongside other activities, the Indian government is optimistic about its?investments doubling or tripling in the coming years.

Atlassian Corporation recently unveiled a generative AI feature in its cloud software suite developed in collaboration with OpenAI.?More details here.

Stability AI?announced ?StableLM , a suite of open-source large language models available in “alpha” on GitHub and?Hugging Face .?Read:?StableLM Might be What Stability AI Needs to Survive .?

AWS recently announced the details of 14 space tech startups selected for the?2023 AWS Space Accelerator Program .?Check out the names of the startups here.?

A Bengaluru-based insurtech startup?Plum ?recently released PolicyGPT,?an insurance ChatGPT.

Charlie Marsh , former?Spring Discovery ?and?Khan Academy ?staff engineer, announced the launch of his new company,?Astral , to build developer tools for the Python ecosystem.?Read more here.?

Ahead of OpenAI, a group of researchers have?announced ?MiniGPT-4 -an?open-sourced ?model performing complex vision-language tasks like GPT-4.?

After two years of introducing?DINO , a self-supervised vision transformer model, Meta AI?announced the launch of DinoV2.