The third-party cookie isn't going quietly

Hi Futurist,

Google is making a strange turnaround with third-party cookies, and they’re fighting back hard against OpenAI and Apple with their new Gemini releases. However, they still can’t generate as much hype around it as OpenAI did in the last two weeks. Therefore, I am sending you insights, inspiration, and innovation straight to your inbox. Let’s dive into the depths of the digital future together and discover the waves of change shaping our industry.

In this post, we're dipping in:

• 📣 Byte-Sized Breakthroughs: Three (technological) developments in the digital landscape

• 🎙️ MarTech Maestros: one webinar, podcast, or interesting paper I believe you should watch, listen to, or read too

•  🧐 Last but not least: Things that I like

Do you have tips, feedback or ideas? Or just want to give your opinion? Feel free to share it at the bottom of this dip. That's what I'm looking for.

Prefer to listen rather than read?

The link to the podcast is only available to subscribers. If you haven’t subscribe already, I recommend to do so.

📣 Byte-Sized Breakthroughs

'Byte-Sized Breakthroughs' briefly highlights (technological) developments in the digital landscape

Headstory: The third-party cookie isn't going quietly

Google promised us a goodbye to third-party cookies. Then it postponed. Now? They’ve flipped the script. Here’s what’s happening: Google won’t pull the plug on third-party cookies just yet. Instead, they’re handing over the controls to the users.

Yes, you read that right. Users will now have the power to opt-out of tracking. If history is any guide, most probably will. Remember Apple’s App Tracking Transparency? Most users said “No thanks” to tracking. Expect something similar here. The era of easy tracking is slipping away.

Here’s why: Chrome dominates, with 65% of global browser usage. If three-quarters of those users opt out of tracking, that’s nearly half of the global internet population off the radar. Throw Safari and Firefox into the mix, and the picture becomes even grimmer for those still clinging to third-party data.

So, what does this mean for marketers? Choice means uncertainty. And uncertainty is a marketer’s worst nightmare. Don’t bank on the leftover Chrome users. Let’s get real. This is a warning. The days of relying on third-party cookies are numbered, and smart marketers know it. Now’s the time to double down on first-party data. First-party data is your lifeline. It’s your lifeline in an increasingly privacy-conscious world. It’s the data you own, the data users trust you with, and it’s the key to building lasting relationships. But how?

Start with this: audit your current data collection. Identify where third-party cookies are being used and evaluate the potential impact of their decline. Identify these weaknesses before they become vulnerabilities.

Next, build direct relationships with your audience. Create value that makes users want to share their data willingly. The more direct the relationship, the better.

Then, implement server-side tagging. Move away from client-side tracking to server-side solutions. This approach not only enhances data accuracy but also gives you greater control over user data, aligning with privacy regulations.

Finally, integrate your systems. Make sure your CRM, marketing automation, and analytics tools are speaking the same language. When first-party data flows seamlessly between systems, you gain insights that third-party cookies could never provide.

Let me repeat: soon, half of all internet users will be nearly impossible to track. The decline is already visible. The downward trend has already begun. Get ahead of the curve, build a resilient data strategy, and secure your place in the future of digital marketing.

Google's response to OpenAI's Voice Mode

TL;DR

Google’s new Gemini Live feature redefines mobile assistance with AI-powered voice interaction. This advanced conversational tool is integrated deeply into Android, offering real-time dialogue, multitasking, and even predictive suggestions across apps. Gemini Live is now rolling out, challenging Apple and OpenAI with its innovative, AI-first approach.

Read it yourself?

Sentiment

Google has taken a bold step forward, surprising many by outpacing Apple, who now appears to be trailing behind in the AI race. The AI community is buzzing with discussions on how OpenAI will respond to this significant leap. Google’s move is seen as a comeback, shifting public perception and positioning the company as a leader in AI-powered mobile assistance.

My thoughts

Google’s integration of AI across multiple devices—earpods, smartwatches, and the new Pixel 9—signals a new era of daily AI interactions. These tools are designed to make our lives more seamless, from capturing moments in photos and videos to managing our schedules and even predicting our needs before we articulate them. I believe these innovations will not only make life easier and more enjoyable but also help us embrace AI as a friend, not an enemy. And that’s a shift we need.

Prompt caching lowers costs and latency for Claude API

TL;DR

Anthropic has introduced prompt caching for its Claude API, dramatically reducing costs by up to 90% and latency by up to 85% for large prompts. This feature is now available in public beta for Claude 3.5 Sonnet and Claude 3 Haiku, with Claude 3 Opus support coming soon. Prompt caching enables developers to retain context between API calls, streamlining processes like long conversations, code assistance, and document processing.

Read it yourself?

Sentiment

The introduction of prompt caching has been met with enthusiasm, particularly among developers looking to optimize AI-driven processes. The drastic drop in costs, coupled with improved latency, has caught the attention of many. The ability to maintain high output quality while slashing expenses is a game-changer for businesses, and this innovation is being hailed as a significant leap forward in making AI more accessible and cost-effective.

My thoughts

By reducing the cost of using AI models like Claude, companies can now deliver powerful AI-driven experiences without the prohibitive expenses that previously limited adoption. Lower costs, combined with maintained or improved performance, could make advanced AI solutions more common in everyday applications, benefiting both developers and end users. s more AI models adopt similar features, we’re likely to see a wave of innovation in the industry.

More byte-sized breakthroughs:

• OpenAI introduces structured outputs and quietly boosts GPT-4o
  OpenAI introduced structured outputs to the API—drawing comparisons to Claude's Artifacts. The update promises 100% reliable schema matching, a 50% reduction in input costs, and a 33% cut in output costs. They also improved the model. Online buzz speculates that this update is a scaled-down version of "Project Strawberry" (formerly Q*), hinted at by Sam Altman himself in a recent tweet.

• xAI releases their latest AI model: Grok-2
  The Grok-2 model, now in beta on 𝕏, outperforms GPT-4 Turbo and Claude 3.5 Sonnet in key benchmarks. With access to real-time information and the ability to generate unrestricted images, Grok-2 embraces a broader spectrum of free speech. With both the full version and its streamlined counterpart, Grok-2 mini, xAI is positioned to revolutionize real-time interactions and multimodal understanding.

• Flux.1 sets the bar for text-to-image generation
  FLUX.1 is breaking new ground in generative AI, offering an open-source suite of text-to-image models that are already making waves in platforms like Grok and Freepik. Known for its exceptional ability to render text within images. Soon, the company behind Flux, Black Forest Labs, will take this innovation even further with the release of a new video generation model.

🎙️ MarTech Maestros

In 'MarTech Maestros,' I share one webinar, podcast, or interesting paper I've encountered, which I believe you should watch, listen to, or read too.

The State of AI at Work

As AI adoption surges, many organizations are unprepared for its impact. With 52% of knowledge workers using generative AI weekly—a 44% increase in just nine months—businesses are confronted with a significant challenge. This report explores the various applications of AI across industries, highlighting impressive productivity gains while also exposing substantial gaps in AI literacy, strategic planning, and adoption. Discover how leading AI organizations are setting the standard, and learn how to position your company for success.

 🧐 Last but not least: Things that I like

• Google launches sentence-level verification for Gemini: Google introduces a new feature to verify the accuracy of Gemini's responses through direct source attribution via Google Search.

• Google DeepMind's AI robot plays table tennis: Google DeepMind introduces an AI-powered robot that achieves amateur human-level performance in table tennis.

• Google's Imagen 3 AI model now available: Google's Imagen 3 AI model is now generally available in the US with improved prompt adherence, including negative prompts.

• Google ruled to have illegally maintained monopoly: A US judge ruled that Google acted illegally to crush competition and maintain its monopoly on online search and advertising.

• Google Gemini 1.5 Pro tops Chatbot Arena: Google's Gemini 1.5 Pro surpasses GPT-4o/Claude-3.5, achieving top scores in multi-lingual tasks, Math, and Coding in Chatbot Arena. The party was short-lived. In the meantime, Gemini 1.5 has already been dethroned from the number 1 spot.

• Character AI CEO Noam Shazeer returns to Google: While Elon Musk has considered buying Character AI, their founders joining the Google’s DeepMind team. Also, Character AI signs a non-exclusive tech agreement with Google.

• Elon Musk sues OpenAI for prioritizing profits: Elon Musk has filed a lawsuit against OpenAI, claiming it prioritized profits over its founding mission to benefit humanity.

• Microsoft lists OpenAI as competitor in AI and search: Microsoft now considers OpenAI a competitor in AI and search, despite their $13 billion investment and long-term partnership.

• OpenAI is making a $60 million bet on an AI-powered device: The ChatGPT maker is leading a $60 million Series B funding round for Opal, previously known as Opal Camera, according to two people involved.

• OpenAI's cautious approach to AI text detection: OpenAI is developing a text watermarking tool to detect ChatGPT-generated writing but is cautious about its release due to potential risks and impacts.

• ChatGPT Free users can create images with DALL·E 3: OpenAI is rolling out the ability for ChatGPT Free users to create up to two images per day using DALL·E 3.

• OpenAI releases SWE-bench Verified: OpenAI introduces SWE-bench Verified, a human-validated subset of SWE-bench, to better evaluate AI models' software engineering capabilities.

• NVIDIA introduces James, an interactive digital human: James uses NVIDIA technologies to provide natural and immersive responses about their products. You can talk to him here.

• NVIDIA's AI training practices raise legal concerns: Nvidia used videos from YouTube and other sources to train AI, raising legal and ethical questions about copyright compliance.

• NVIDIA faces DOJ antitrust probes over AI dominance: The US DOJ is investigating NVIDIA’s AI business for potential antitrust violations, including its acquisition of Run:ai and market dominance tactics.

• Figure 02 robot unveiled: Figure introduces its new humanoid robot, Figure 02, featuring a complete hardware and software redesign and tested at BMW's Spartanburg plant.

• Runway introduces GVFX for all footage: Runway announces the addition of generative visual effects (GVFX) to any footage, from phone videos to high-quality cinematic action.

• Testing Runway Gen-3's new first and last frame feature: Gen-3 introduces a feature allowing longer and more complex video clips by using the frame as the midpoint.

• Reddit to test AI-powered search result pages: Reddit is planning to test new search result pages powered by AI.

• Amazon upgrades its AI image generator: Amazon has released an upgraded version of its image-generating AI model, Titan Image Generator, with several new capabilities.

• YouTuber files class action suit over OpenAI’s scrape of creators’ transcripts: A YouTube creator alleges OpenAI used millions of YouTube transcripts to train AI models without consent or compensation, seeking over $5 million in damages.

• Instagram embraces photo dumps with 20-photo carousels: Instagram now allows users to add up to 20 photos or videos in a single carousel post, doubling the previous limit.

• Groq lands $640M to challenge Nvidia: AI chip startup Groq raises $640 million to compete with Nvidia in the generative AI hardware market.

• AI-generated song reaches German Top 100 charts: A satirical AI-generated song about teenage gangsters, "Talahons," has entered the German Top 100 charts. Listen here.

• GitHub introduces AI model playground: GitHub launches GitHub Models, enabling developers to experiment with AI models directly on the platform, from playground to production deployment via Azure.

• Introducing Stable 3D: Stability AI's latest breakthrough in 3D asset generation technology, transforming a single image into high-quality 3D assets in just seconds.

• AI music startup Suno responds to RIAA laws: Suno claims its use of copyrighted music for training AI models is 'fair use' in response to the RIAA's lawsuit.

• Copyright Office urges Congress to outlaw AI impersonation: The U.S. Copyright Office calls for urgent legislation to combat unauthorized AI-generated digital replicas, citing threats to reputations and livelihoods.

• World's most capable AI software engineer: Achieved 30.08% on SWE-Bench, surpassing Amazon and Cognition, trained to think and behave like a human SWE.

• Mistral AI announces new advancements: Mistral AI introduces model customization, alpha release of Agents, and stable version of their client SDK to enhance software development.

• MultiOn's new Agent Q release: MultiOn's new Agent Q, possibly powered by OpenAI's Strawberry upgrade, combines advanced AI techniques for improved learning from both successful and unsuccessful actions.

• The AI Scientist: Sakana AI introduces the world's first AI system for automating scientific research and open-ended discovery, capable of generating and reviewing entire research papers.

• Artists' lawsuit against generative AI makers can go forward: A judge has allowed a class action lawsuit by artists against Stability, Runway, and DeviantArt for allegedly using copyrighted works to train their AIs to proceed.

• Neura shows off humanoid robot 4NE-1: German robotics manufacturer Neura released a video of its humanoid robot 4NE-1 performing various activities.

• The EU's AI Act is now in force: The European Union's risk-based regulation for applications of artificial intelligence has come into force starting from August 1st.

• EU closes TikTok Lite case after rewards feature withdrawal: TikTok Lite's rewards feature investigation ends as TikTok commits to permanently removing the mechanism from the EU market.

• Hollywood actors’ union SAG-AFTRA sets AI audio agreement with Narrativ: Allowing actors to license their digital voice likeness on the platform.

This was it. Our eighteenth digital dip together. It might seem like a lot, but remember; this wasn't even everything that happened in the past few weeks. This was just a fraction.

Of course, it can be confusing at times. One moment we're talking about a data strategy for AI applications, the next for third-party data. But it's essentially the same. It all comes down to centralizing first-party data. And the roadmap to get there? I’m happy to help with that. Just hit reply, and let’s schedule a meeting.

Looking forward to what tomorrow brings! ▽

-Wesley