Google's Redemption? Unveiling Gemma 2 and Gemini Upgrades
It's been a quiet Summer for Google since May's I/O event — but the tech giant just released an impressive double-feature blockbuster. With the launch of Gemma 2 and several new Gemini upgrades, is Google about to steal the AI spotlight once again? Let's investigate...
In today’s AI news:
Read time: 4 minutes
The News:
Google just launched Gemma 2, the next generation of its open lightweight AI model series — alongside new upgrades to its Gemini 1.5 Pro model including a 2M token context window and enhanced coding capabilities.
The details:
Why it matters: Google’s Gemma upgrades send a new open-source powerhouse to the top of the rankings while running efficiently (and affordably) on a single GPU. Not to be overshadowed is Gemini’s 2M context window release, which opens up a mind-blowing number of new capabilities to users.
The News:
TIME just announced a new multi-year licensing agreement and partnership with OpenAI, infusing the media giant’s library of content into ChatGPT, AI model training, and other products.
The details:
Why it matters: ‘If you can’t beat em, join em’ seems to be the theme of the current state of AI and journalism — and it’s getting lonelier on the side of publishers still trying to fight against the current. With the recent news of MTV’s archives being deleted, maybe AI models can not only inform but also act as a preservation tool for years of content currently fading from the Internet.
The News:
OpenAI just published new research exploring how to improve model training using AI ‘critics’ that help human trainers spot errors in AI outputs more effectively.
The details:
Why it matters: As OpenAI states in the research paper, both LLMs and AI critics will continue to scale — human intelligence will not. The limitations of our own puny mortal brains look to be no match for evaluating systems that are continuing to soar, meaning AI is quickly becoming its own trainer.
Character AI
Launched Character Calls, a free feature enabling users to have two-way voice conversations with AI characters across multiple languages on the platform’s mobile app.
University of Tokyo researchers
Developed a new technique to bind living human skin to robotic faces, potentially enabling more lifelike androids in addition to other medical applications.
Meta
Meta is set to begin testing its creator-made AI chatbots on Instagram in the U.S., with a broader rollout expected by August.
Eureka Health
Introduced Eureka, an ‘AI doctor’ offering personalized care plans, lab orders, and medication adjustments under physician supervision with a specialty in thyroid and diabetes treatments.
New York University researchers
Published a new paper on MLLMs and vision-centric approaches, also introducing a new Cambrian-1 model family and a new benchmark to measure vision-related tasks.
Thank you for reading our newsletter! If you want to stay two steps ahead of the competition, subscribe to this newsletter. If you want to leave your competition in the past, hop on a quick, complimentary, no-obligation call with our team to explore our consulting and custom development services.
We've proudly worked with over 400+ companies to revolutionize their business with AI, and our team of 3,500+ developers, engineers, consultants, and experts are more than ready to help you take advantage of all the latest technologies.
