Hello There:

We are introducing a new feature you will like. ❤️ Trendscape 2024 is a list of the top trends that have the internet buzzing these days. 📊 I am only sharing this soft launch with you, my faithful readers. 😻

The No. 1 trend on everyone’s minds is privacy. Given the record-breaking 26 billion records spilled out into the wild in 2024, including more than 200 million social security numbers, that’s not surprising.

🤔 Think about this:

There were many others. Coming in next was Facebook, whose 3.3 billion users helped propel it to second place. It was, however, not just cute photos of cats and dogs that boosted its profile, but innovations like Meta AI and Ray-Ban Meta smartglasses that put artificial intelligence in reach of the masses.

Two trends loomed large in 2024. The economy and China, joined at the hip like belligerent Siamese twins, with one threatening the other. To put it bluntly, China (No. 23) and Russia (No. 67) are our two biggest threats in 2025, aside from climate change (No. 47).

In every issue, we will feature a Trendscape trend (see below).

We hope you enjoy the resource.

01. Privacy

Steel cuffs clicked shut around Pavel Durov’s wrists, the cold bite of metal starkly contrasting with the Parisian summer heat. The irony wasn’t lost on him. Here was the champion of digital privacy being detained by a government demanding more transparency.

In August 2024, Durov, the Russian-born founder of Telegram, was arrested and charged in France for failure to cooperate with law enforcement and complicity in crimes committed on the service, including the distribution of child sexual abuse material, drug trafficking and fraud.

French authorities were adamant. Telegram, his encrypted messaging app, had become a haven for criminals. Drug deals, terrorist plots, and child exploitation — all thriving within the secure confines of his creation. Yet, Pavel remained steadfast in his belief that privacy was an inalienable right and that communicating without fear of surveillance was paramount in an increasingly Orwellian world.

News of his arrest rippled across the globe. Privacy advocates rallied, tech titans voiced their concerns, and a palpable sense of unease settled over the internet. The incident spotlighted the ongoing tug-of-war between governments demanding access and individuals yearning for privacy.

In the quiet of his cell, Pavel realized the world had changed drastically since he launched Telegram. Today, an obsession with privacy permeates society. Every click, every message, every search is now a commodity, scrutinized and monetized. His app, once a niche tool for the tech-savvy, had become a lifeline for millions.

After four days of detention, Pavel was released after posting bail of €5 million with the condition that he report to a police station twice a week and remain in France. Pavel walked out of the courthouse and was ushered into a waiting car by a bodyguard. But he knew the fight was far from over. The delicate balance between privacy and security would always be a precarious one.

Continue reading on Ubertrends…

2024 AI Trends

In the annals of artificial intelligence history, 2024 will be remembered as a year in which AI advanced at a torrid pace. Three trends stood out last year: agentic, speech synthesis and workflow.

Agentic

Agentic is a new-age term that describes the emerging trend of “AI Agents.” These intelligent systems are moving beyond simple question-answering, towards actively performing tasks, automating workflows, and interacting with software and data on our behalf. The shift is fueled by advancements in model architecture and a growing focus on reasoning capabilities.

Agents are applications that can act, react, and interact with both humans and other agents alike in achieving goals. Anthropic may have fired the first shot on Oct. 22, 2024, when it announced in true Silicon Valley speak that it had developed a “next-gen algorithm for AI self-teaching.” Put plainly, an AI assistant able to perform research, answer emails, and handle other back-office jobs on its own.

The company claims to have an improved Claude 3.5 Sonnet model that can comprehend and communicate with any desktop application. The model can mimic keystrokes, button clicks, and mouse movements through a new “Computer Use” API, still in beta, simulating a human sitting at a computer.

In June, Anthropic introduced Artifacts, which let users view, edit, and build in a real-time side panel. This feature was quickly imitated by OpenAI on December 10 when it launched Canvas and by Google in Gemini 2.0 Stream Realtime on December 11.

Those launch dates underscore the rapid advancements in large language models that took place in 2024. Google’s Gemini 2.0 Stream Realtime is said to include the same capabilities as ChatGPT Pro, but instead of costing $200/mo., it’s free. Forgive us for almost forgetting that Gemini was called Bard in February.

Almost as popular as the big three, Perplexity was able to garner a lot of ink due to its formidable research capabilities, which continue to outpace the big three. But like all other models, every fact Perplexity cites needs to be double-checked due to rampant hallucinations. And that remains the biggest shortcoming of generative AI.

The potential for agentic applications to be more proactive in problem-solving, however, is what has attracted widespread interest. In January 2025, OpenAI showed just such an application when it debuted its task-oriented ChatGPT Reminders:

At press time, TechCrunch reported that OpenAI’s agent tool, called Operator, might be nearing release. Operator can take control of your computer and perform actions on your behalf.

To gauge just how far and fast we have advanced, take the A.R.C.-A.G.I. test, for instance, which was designed to compare the fluid intelligence of humans and AI models. In mid-2024, the highest score any AI system had posted was 5%. By the end of the year, the most powerful OpenAI o3 model scored 88%.

Speech Synthesis

The most remarkable speech generation launch came in September 2024 when Google introduced a feature of NotebookLM called Audio Overview that could generate a podcast from any text file uploaded. TechRadar called it “stunningly real.” Judge for yourself:

Workflow

That brings us to the holy grail of generative AI, which is the promise of workflow improvement. After all, the world’s 1 billion information workers need all the help they can get.

The famously easy chatbot interface, which in its early form resembled Google with just a single input field, has quickly become much more complex. With multi-modal engine choices, overlapping menu structures filled with undecipherable icons, and countless input methods — including text, voice, video, screen sharing, and file uploads — the user interface and experience are becoming increasingly paramount.

That’s why we have been following apps that promise to improve workflow with much interest. The first arrivals are not without flaws. The first to land our radar screen was a tool called MindStudio.

It promised to help users rapidly create custom chatbots, which is what just about every business wants to do. A survey by The Verge in early 2024 found that 48% of respondents were interested in building their own AI tools. While responses were likely collected from an unrepresentative sample, many said they used Snapchat, the survey does reveal directionally what the market is interested in.

After grappling with MindStudio, which was better suited for developers, we encountered an Irish startup, AirOps, which promises a breakthrough answer for workflow automation. However, AirOps requires intimate knowledge of JSON, which limits its appeal to SEO shops.

The latest workflow tool we investigated is a startup from Vancouver, Canada, called Gumloop. It offers the slickest interface of the three. Unfortunately, our experience suggests that this tool is not ready for primetime, with key interface pieces missing in their tutorials.

While these workflow solutions all offer integrated chatbots that promise to help improve the journey, our experience is that they are of little help.

This is a challenge the entire industry faces. The hype surrounding AI has inflated user expectations to such an extent that disappointment is certain to follow.

We found the same to be true of the new wave of coding assistants. At first, both Replit Agent and Bolt.new amazed us, but once the novelty wore off, we didn’t get much done because both models kept making annoying mistakes. Replit Agent, for example, was incapable of coding a simple user registration and log-in process, something every SaaS app developer needs.

Despite these challenges, we look forward to telling you in January 2026 that we have had nothing but awesome experiences with applications that will be hot then. 😉

Future of Work

The not-ready-for-primetime nature of the initial wave of applications bodes well in the short term for the future of work. Here are some key findings:

• AI and Growth: Where Do We Stand? – One recent study by a pair of French trend analysts found that the impact of AI on labor demand was positive even for occupations that are often classified as vulnerable to automation, such as accounting, telemarketing, and secretarial work.

• The World Needs a Pro-Human AI Agenda – Those French analysts cite Daron Acemoglu, who believes “we should not expect much more than about 5% of what humans do to be replaced by AI over the next decade.” Yet, in that same article, Acemoglu comments: “There is no evidence yet of AI delivering revolutionary productivity benefits.” He clearly is not one of those legions of users who are causing OpenAI to lose money on its $200-per-month ChatGPT Pro plan because it has no productivity benefits.

• Future of Jobs Report 2025 – Respondents to a World Economic Forum study saw things differently. Fully 60% of employers expect “broadening digital access” to transform their business by 2030, with 41% planning staff reductions that year due to “skills obsolescence” — likely due to AI. That finding corresponds with a study by Rasmussen that revealed that 56% of Americans believe that within the next 25 years, most jobs in America will be done by robots or computers.

• Philippines call centers – Rest of the World reports that AI is making Philippines call centers more efficient, for better or worse. One worker tells the publication that AI copilots have doubled his call-handling volume. Industry estimates suggest that 300,000 Filipino BPO employees could be out of a job due to AI in the next five years, with only 100,000 new jobs created in roles such areas as data curation. That means that 11% of the industry’s 1.84 million workers would be unemployed.

Happy New Year! 🥳

The Jacobyte is Back! 😂 

Meet The Jacobyte — an irreverent and sharp-tongued columnist who dominated the digital marketing industry’s gossip mill in the late ’90s and early aughts, delivering an unfiltered mix of innuendo, rumors, and biting satire. Back to shake things up, Jaco returns with his signature skewed perspective, exposing the absurdities, hypocrisies, and hidden truths of today’s digital infested world. Brace yourself for the wit, the grit, and the muckraking you didn’t know you missed. — Ed.

😜 Remember when the excuse was the dog ate my homework? Well, some teachers may wish the old days were back, because, according to Pew Research, 26% of U.S. teens have used ChatGPT for schoolwork. Hey, little boys and girls, Jaco can’t decide if you’re smart learning to use a chatbot at an early age or just growing up ignorant by not applying yourself in earnest.

🤦 Speaking of ignoramuses. Did you hear that Google’s NotebookLM chatbots had to be taught not to be annoyed when humans, using a new feature called “Interactive Mode,” which lets users “call in” to a podcast and ask questions, essentially interrupting the AI hosts as they talk?

When the feature first rolled out in December, the AI hosts uttered such passive-aggressive comments as “I was getting to that” or “As I was about to say,” which felt “oddly adversarial.” LOL, you mean bots have to be taught not to act like humans? The story made Jaco hallucinate, inducing a ketamine “Her” moment. 🤭

🤖 Shootout at the robot coral! In a scene right out of “I, Robot,” a Serve delivery robot collided with a Waymo robotaxi in Los Angeles. The Serve bot, who will call Sonny, was crossing a street in West Hollywood at night and navigating the curb ramp, when Sonny was hit by a big, bad Waymo. If you think this is like real-life science fiction, you are right, moose breath! Watch the video below.

❤️ Remember that the Chairman of the Board sang, “L.A. Is My Lady,” and she needs our help. The Eaton and Palisades fires are just 89% and 63% contained, and tens of thousands of people are in need of housing, food and clothes. While the American Red Cross is everyone’s top charity, Shelly Palmer has pulled together an excellent resource, “Top-Rated For Los Angeles Wildfire Relief.” Thanks, Shelly! 🥲

🤩 Speaking of smokin’, Jaco has to laud Houston’s Brandy Deason. You may not remember the name, but she was the one who came up with the ingenious idea of using 12 Apple AirTags to find out where her recycling ended up. She tracked most of her recycling to massive piles of plastic at the Wright Waste Management facility. Her brilliant deed led to her being called ”The James Bond of Plastic Recycling.”

Inspired, Jaco is now feverishly trying to figure out where to stuff Apple AirTags in order to be called the “Austin Powers of Generative AI.” Oh, behave, you International Man of Mystery! 😂

Sayonara, boys and girls. 👏

Send your rumors, tips and contributions to: [email protected].