Welcome to Part II of Edition No. 63 of my weekly newsletter, providing practical analysis in the world of digital content strategy.
Friday Five
I. With Gemini, Google Says it’s Surpassed ChatGPT4
II. Readers Want Transparency Around AI…At a (Fake) Cost
III. Google Signals To Be Removed from GA4 Reporting Identity
IV. Meta’s Latest AI Investments
V. More AI Tools for LinkedIn Premium Members
VI. Other Important Updates
I. With Gemini, Google Says it’s Surpassed ChatGPT4
Google's DeepMind team has introduced its latest version of AI, called Gemini.
It's designed to work just as well on big company servers as it does on your phone. When tested, Gemini was better at understanding languages and writing code than even human experts, showing that AI can now solve problems in ways we hadn't seen before.
“It was built from the ground up to be multimodal, which means it can generalize and seamlessly understand, operate across and combine different types of information including text, code, audio, image and video.”
It has demonstrated superior performance, outclassing previous models, including GPT-4, in language understanding and coding benchmarks, showcasing exceptional reasoning capabilities.
Google's responsible AI ethos is embedded in Gemini's development, emphasizing safety evaluations, bias and toxicity assessments, and collaborative efforts with external experts for a more inclusive AI.
There are three versions: Ultra, Pro and Nano, from most powerful to least.
🛠 Why does this matter?
Where can you access it? “Starting today, Bard will use a fine-tuned version of Gemini Pro for more advanced reasoning, planning, understanding and more.” Pixel phones will also have the Nano version.
Google says its approach to Gemini's development — prioritizing responsible and ethical AI — aims to ensure that such powerful models are deployed with a focus on safety, accountability and societal benefit.
In comparison to ChatGPT-4 (see below), Gemini's superior performance in benchmarks is significant. This implies that Gemini could offer more precise, nuanced and reliable outputs, which is crucial for applications requiring complex reasoning and decision-making.
The flip side of this? The existential arms race that concerns a lot of people – including those on the inside of the AI tool creation – will only heat up. Is there anything that could motivate OpenAI more than Google releasing this side-by-side comparison claiming their tools is superior?
•
II. Readers Want Transparency Around AI…At a (Fake) Cost
“An overwhelming majority of readers would like news publishers to tell them when AI has shaped the news coverage they’re seeing. But, new research finds, news outlets pay a price when they disclose using generative AI.”
“More than three-quarters of U.S. adults think news articles written by AI would be ‘a bad thing.’”
🛠 Why does this matter?
There a few questions I would like to ask those same readers who are anti-AI for content creation:
Is accuracy the top priority for your news content?
Should human editors review every word of a piece of content before it’s published?
Do you prefer to have access to news as quickly as possible so long as it’s accurate?
If the answer to all those is “Yes” and AI is helpful in achieving those outcomes, then I think it’s contradictory to say you aren’t in favor of creators using AI tools to produce content.
From Nieman:
“On an 11-point trust scale, survey respondents who saw the news stories labeled as AI-generated rated the mock news organization roughly half a point lower than those shown the article without the label — a statistically significant difference.
The respondents, interestingly, did not evaluate the content of the news article labeled as AI-generated as less accurate or more biased.”
I rest my case.
•
III. Google Signals To Be Removed from GA4 Reporting Identity
In addition to the soon-removed Google Signals, there are three other “Identity spaces”: User-ID, Device ID and Modeling.
Modeling is basically GA4 using AI to determine user identities.
🛠 Why does this matter?
You can already do this manually in your GA4 admin. This will make it so that you don’t even have the option to include Google Signals in your reporting identity.
Since Google Signals – which you must activate in your account in order to use (it's off by default) – is used to collect personal data such as gender, age and interests, I imagine this to be a legal-driven decision. Especially in the light of ever-tightening EU restrictions on user data.
Most users don’t even know you can adjust user identities. One good reason to do it? Not using Google Signals in your reporting identity drastically reduces thresholding, meaning you’ll have more accurate data.
•
IV. Meta’s Latest AI Investments
Meta is implementing more than 20 new AI-driven features across Facebook, Instagram, Messenger and WhatsApp.
New features include improved search functions, enhanced social discovery and advanced advertising and business messaging capabilities.
Meta AI, the company's virtual assistant, is evolving with enhanced messaging features and powering smart capabilities in the background.
A notable addition is invisible watermarking in AI-generated images for better transparency and traceability.
Meta AI has been integrated into various applications, offering functionalities like photorealistic image generation, AI chat options, and more.
The company has also introduced a new feature called “reimagine” for interactive image creation in group chats and integrated Reels into Meta AI chats.
🛠 Why does this matter?
These advancements indicate Meta's significant investment in AI to enhance user experience and functionality across its platforms.
The introduction of invisible watermarking addresses concerns about the authenticity and origin of AI-generated content, promoting transparency.
The integration of AI into social media activities, like creating and sharing images or browsing Reels, illustrates a shift towards more interactive and AI-driven content consumption.
•
V. More AI Tools for LinkedIn Premium Members
“Searching for a job? Tap into new AI-powered advice on any job posting, to learn more about a role or company, understand your fit for the role, and get personalized tips to make your application stand out.”
“Scrolling your LinkedIn feed? Get personalized AI-powered tips and advice on posts, go deeper on a topic, and see how to apply the learnings to your job. See personalized takeaways on this post by clicking on the suggested prompts with the gold sparkle icon below.”
“…alongside existing features like an AI writing assistant to enhance your profile and help you draft messages.”
🛠 Why does this matter?
LinkedIn may not be building its own chatbot (yet) like Google, X or OpenAI, but it’s not going to get left in the dust, either.
And since Microsoft, which owns LinkedIn, owns 49 percent of OpenAI, expect a lot more in the way of AI tools.
•
VI. Other Important Updates
•
Any other important updates this week? Let me know in the comments.
I self-edit this newsletter, so please don’t be shy about emailing me if you spot a mistake.
See you next week for a practical tip and an in-depth analysis in the world of digital content strategy.