Last week, Databricks made a resounding entry into the open-source large language model (LLM) arena with the release of DBRX. This move didn't come as a surprise, as DBRX was built by the Mosaic research team, which Databricks acquired last summer through the acquisition of MosaicML.
Databricks' core offering revolves around providing a powerful analytics platform that enables its customers to leverage their organizational data effectively. However, generative AI capabilities were previously missing from Databricks' arsenal. With the MosaicML acquisition, it was expected that Databricks would enhance its data analytics offering to incorporate generative AI capabilities, and that's precisely what they've achieved with the DBRX release.
By seamlessly integrating DBRX into their platform, non-technical companies gain the ability to harness the power of their own data, unlocking their industry-specific insights that were previously inaccessible without substantial investments in developing in-house artificial intelligence and machine learning expertise. This democratization of generative AI capabilities empowers businesses across sectors, enabling them to leverage cutting-edge technology without the need for specialized skills within their workforce.
One key advantage I see with DBRX, that might propel its adoption in the enterprise landscape, is that it’s backed by an established professional services company like Databricks which is currently lacking for other open-source LLMs.
Key updates of the week
Apple's researchers have unveiled a novel AI system dubbed ReALM in a fresh research paper, capable of comprehending on-screen tasks, conversational contexts, and underlying background processes.
Researchers from Meta, Cisco, and MIT have unveiled findings, in a recent paper, that suggest large language models (LLMs) can undergo substantial pruning, with up to 40%-50% of their layers being removed, while maintaining minimal impact on their accuracy and performance.
X.ai has unveiled Grok-1.5, an upgraded AI model boasting improved performance on math and coding benchmarks compared to its predecessor.
While AI-powered search platforms such as Perplexity strive to challenge Google's dominance, they still grapple with developing a sustainable search business model.
OpenAI and Microsoft are collaborating on a massive $100 billion data center initiative called "Stargate" planned for launch in 2028.
Samsung has acknowledged the necessity to revamp its voice assistant, Bixby, through the integration of cutting-edge generative AI capabilities.
Insights from analytics firm Similarweb suggest that OpenAI's post-ChatGPT era is unfolding at a pace quicker than widely expected.
Community updates
Share your updates with other readers by telling us what you are working on here.
Your support is vital
Thank you for your continued support of the newsletter and the growth so far. Did you find this issue valuable? If so, there are two ways you can help:
You can also hit the like ❤️ button at the bottom of this email to help support me or share 🔄 this with a friend. It really helps!
Subscribe to a monthly/yearly subscription to support my work. Alternative, make one time contribution if you like to go down that route.