NOTEWORTHY NEWS: OUR TAKE ON THE LATEST AI/GENAI NEWS
Noteworthy AI News for Agencies - Edition #17
By Jeremy Lockhorn , SVP, Creative Technologies & Innovation, 4A's
THE BIG STORY:
Big tech drives AI advancements
Apple kicked off the week with its annual iPhone reveal, announcing new hardware including Apple Watch Series 10, new airpods and the iPhone 16 lineup. Apple Intelligence featured prominently throughout the event, including a new “Visual Intelligence” feature, which is one of several features leveraging 3rd party services. Some reviewers are less than impressed with Apple Intelligence, but regardless - the rollout is bound to introduce a wide audience to a new set of AI features leveraging hardware and software integration, built right in at the OS level to your most personal device.
OpenAI, which is of course one of the other 3rd parties enabling AI features on iPhone, also made headlines this week. The Information reported on Tuesday that OpenAI’s next model, the long-rumored Strawberry, was due to be released in the next couple of weeks. This new model has been rumored to focus on reasoning, taking time to “think” before responding - making it potentially a better performer for complex queries that may have taken multiple prompts for ChatGPT-4o to accomplish.?
Seems like The Information was onto something. Yesterday, a blog post revealed OpenAI o1-preview. This is actually a series of models, including the main o1 LLM, and a smaller version called o1-mini. Both models have a renewed focus on safety and security, following on the heels of OpenAI’s recently formalized agreements with the U.S. and U.K. AI Safety Institutes. o1 apparently “performs similarly to PhD students on challenging benchmark tasks in physics, chemistry, and biology.”?
More updates are in the works of course, but this release promises to improve ChatGPT’s ability to solve complex problems, which could prove useful for common agency challenges like budget allocation, detailed audience segmentation, deep personalization, campaign optimization, and more. It is not a replacement for ChatGP- 4o, which will likely continue to perform better for some use cases. o1 is not multimodal, for example. And it is more expensive for API access. The advanced reasoning capabilities may also require different prompting strategies to unlock full potential.?
OTHER INTERESTING AI NEWS FOR AGENCIES:
RECENT CASE STUDIES, EXAMPLES AND USE CASES:
We'll see you next week with another edition of #4AsNoteworthyAINews.
Visit our AI/GENAI Hub for the latest guidance from the 4A's.
领英推荐
Missed previous editions of Noteworthy News? Catch up here:
#AI #GenAI #ArtificialIntelligence #EmergingTechnologies