Google MedLM is a family of core templates optimized for the healthcare industry on Google. … [+]
Artificial intelligence (AI) is evolving. But let’s not forget where we come from. The first notions of pseudo-sentient intelligence that emerged from the mainframe laboratories of the 1950s were perhaps too embryonic for the processing and storage power of the time. While they may have given way to the “cinematic AI” of the 1980s, we only started to see real progress in the post-millennial years and IBM Watson gained its fair share ( and more) attention in this space.
Of course, AI is changing again and it’s not hard to see why. The rise of generative AI (gen-AI) from large linguistic models (LLM) running on vector databases has not been forgotten in technology news all year long.
More precise and refined AI tools
But as we enter a new year and perhaps some of the furor and hype fades, what happens next with AI is a matter of refinement and tooling, that is- that is, we will now create more precise language models that align with industry, task or function. specific jobs…and our current focus is on creating better tools for software application development professionals to integrate new strains of AI into our applications.
Google capped a year of hysteria around the AI generation with the launch of its Gemini Large Language Model.
Before we look at how Google is positioning Gemini to reflect current trends, let’s stop for a nanosecond and remember what we just said here, which is that the IT industry isn’t talking about an engine or a higher level AI model, tech glitter is. I’m not focusing on a new AI-enriched app that will order you a new pint of milk when the RFID-tagged carton in your fridge says an expiration date…and we’re not talking about a new AI widget that will surface. on our smartphones. Instead, we are excited about a new data science approach at the lower substrate level that will propagate upward to give us better AI. As we said, AI is evolving.
Fanfares aside, what we can see here is that Google is very much reflecting the need to refine and refine AI at this point. Technologists want AI tools that can ingest any type of data and perform in the widest variety of post-deployment scenarios. Google knows this and has built Gemini to be “multimodal” and capable of ingesting information in the form of text, but also in the form of images, audio and video.
Gemini triplets
While we normally think of Gemini couples as twins astrological terms at least this Gemini was shaped and sized like a triple pack. By creating different versions of Gemini, Google says it will “work effectively” on everything from data center-level cloud deployments to mobile devices. To enable enterprise software application developers to build and scale with AI, Gemini 1.0 has been optimized in three different sizes:
- Gemini Ultra: The largest and most powerful model for very complex tasks.
- Gemini Pro: the best-suited model to accommodate a wide range of tasks. Calling him versatile might not do him any favors, but you get the point.
- Gemini Nano: As the diminutive name suggests, the most efficient model for on-device tasks.
With the real-world interests of software developers at the forefront, the company now confirms that Gemini Pro is available through the Gemini API for Google AI Studio developers, the company’s development environment designed to enable programmers to integrate Gemini models via an application programming interface (API). and develop prompts as they create code to build generative AI applications. It is also available to businesses through Google Cloud’s Vertex AI platform, as explained here.
Why is Gemini available via both routes? The API option through AI Studio is a free web development tool designed to encourage usage and generate interest. Google says that when coders are ready for a fully managed AI platform, they can move their AI Studio code to Vertex AI for additional customization and Google Cloud features, at a cost, as we know, There is no such thing as a free AI lunch.
Shaping AI for health
If the trend toward shaping and refining (and we can generally take scale as a given) AI is currently born out of Google’s work with these tools, we can see it in the introduction of MedLM, a family of models core refined for the healthcare industry, available to Google Cloud customers in the US through Vertex AI, this technology will be more widely available next year.
The company is keen to show a friendly face as it attempts to encourage coders to get involved with its AI technologies by providing additional tools and support. According to Google’s AI blog, “Duet AI for Developers is now available to everyone. This permanent Google Cloud contributor provides AI-powered code and chat support to help users build applications in their favorite code editor and software development lifecycle tools. It also streamlines running applications on Google Cloud, and Duet AI for Developers provides businesses with built-in support for privacy, security, and compliance requirements. We will be integrating Gemini into our Duet AI portfolio over the coming weeks.
What happens next, globally
Although Google has reflected (some would say motivated, others would say followed) trends in the AI industry as a whole and has worked to refine and shape AI from the way it ingests information to how it can be applied, there are (obviously) still challenges ahead. Although many of these technologies are available in all territories, Google is rolling them out first in the United States, then in Europe (and the rest of the world). So in terms of international deployment factors and perhaps governance, there is a broader question for the future.
We’ve engaged the medical industry here, work is also underway to deliver Google Duet AI into the security operations (SecOps) space and make generative AI generally available to defenders in a SecOps platform unified. This is great for security teams, but there are many other technology engineers in a) the operations team and b) the broader IT department who will want to participate in the generative AI movement and be able to work simultaneously (software parallelism pun) with their colleagues.
Artificial intelligence is evolving and it will continue to do so – although many believe this year of generative AI stands out as a pivotal moment in time – let’s hope developers have the right tools and we’re not hallucinating.