SODP Dispatch - 16 January 2025

Tech Companies Are Turning To ‘Synthetic Data’ To Train AI Models – But There’s a Hidden Cost

Tech companies depend on data – real or synthetic – to build, train and refine generative AI models such as ChatGPT. The quality of this data is crucial.

Real data refers to text, video and images created by humans. Synthetic data is artificially created or generated by algorithms.

In theory, synthetic data offers a cost-effective and faster solution for training AI models. It also addresses privacy concerns and ethical issues, particularly with sensitive personal information like health data.

But although synthetic data offers promising solutions, it is not without its challenges.

📈 Media trends report

Reuters Institute has published its Journalism, media, and technology trends and predictions 2025 report.
See more ▸

🕵🏼‍♂️ Real-Time News Monitoring

MyPressAlert - currently in beta – is a platform that monitors news broadcasts in real time and matches your topics of interest to spoken words.
See more ▸

By Vahe Arabian

🚀 Boosting first-party data acquisition

First-party data is a critical asset for digital publishers as it provides direct, reliable insights into their audience's behavior, preferences, and engagement patterns. This data is not only more accurate and privacy-compliant but also empowers publishers to build deeper audience relationships, deliver personalized content, and create targeted monetization opportunities.

In an era of increasing data privacy regulations, leveraging first-party data is essential for publishers to maintain their competitive edge and drive sustainable revenue growth.

Here is where publishers who are seeking to boost first-party data acquisition, consumer trust, and compliance with regulations can get started:

• Progressive Profiling

  Asking for too much data at one go can come across as very invasive. Instead, a better option is to gradually build relationships with your audience while slowly acquiring data by offering innovative content and experiences.
  

• Enforcing Clarity and Transparency

  A lot can go wrong if your teams don't have a good understanding of their responsibilities and limitations about data collection. Proper training and awareness about what to collect, when to collect, and how to use it can help prevent future headaches.
  

• Using a CDP

  A Customer Data Platform can help bring all your audience data across devices and platforms to a single unified interface. For example, The Guardian is exploring CDPs as it is a large publisher that does not impose mandatory sign-in and has millions of anonymous users.

Google quietly updated the News and Discover manual action policies | SEJ

Google has updated its manual actions for Google News and Google Discover a few months back, where they removed the adult-theme and artificial freshening sections, updated the hateful content section and added five new deceptive practices sections.
Read more ▸

Google partners with the AP to provide fresh results for the Gemini app | Google Blog

The AP will now deliver a feed of real-time information to help enhance the usefulness of results displayed in the Gemini app.
Read more ▸

Substack rolls out livestreaming for all publishers | TechCrunch

After testing live video features in the fall, newsletter platform Substack announced on Tuesday that it is now making livestreaming available to all publishers. This new feature allows creators to engage with their viewers and collaborate with other publishers.
Read more ▸

Google Shows How To Fix LCP Core Web Vitals | SEJ

Google Chrome's Barry Pollard explained 5 optimization tips for Largest Contentful Paint. Every SEO needs to read this.
Read more ▸

The Conversation is trying to make its academia-fueled model work for local news | NiemanLab

The nonprofit outlet has built a brand on connecting the knowledge of professors and researchers to both the news cycle and a general audience. But could it also work on a sub-national level — in a city or state — and try to make a dent, however small, in the local news crisis?
Read more ▸