Five Trends January 2024

Here are five interesting trends that I’ve been reading about recently.

Synthetic Data

Data is the primary input for AI technology. One big challenge for technology teams is not having enough data to train their AI models.

Amazon recently posted how they used generative AI to overcome this data challenge when building one of their products. The product is called Amazon One. It’s a contactless palm scanning device.

The device’s AI technology was trained on millions of images of palms. However the team didn’t actually have millions of images of palms. The team used generative AI to generate a dataset of millions of palm images, and then used that dataset to train Amazon One. This type of data is called synthetic data.

We used generative AI to create a “palm factory”—producing millions of synthetic images of palms—to train our AI model. The industry term for this computer-generated information is “synthetic data,” new data created by the AI to replicate the breadth and variety of real data as closely as possible. This was pioneering work that happened years before the current generative AI craze started dominating conversations around the world.

About Amazon

It’s an interesting use case where generative AI is used to create data, that data is then used to train AI models that produce more data, and the cycle repeats…

Third Places

Brands are starting to experiment with creating public places being branded as a third place.

Interest in “third places”—public spaces that aren’t home or work where people can gather, hang out or just be—is growing as the line between people’s offices and living rooms grows ever-blurrier in the hybrid-work era. Some brands that don’t usually have anything to do with hospitality say that demand is an opportunity.

Wall Street Journal

The ventures aren’t about creating another meaningful revenue stream or sales pipeline, according to the companies. Rather, they are designed to make a statement about brand values in a physical form that is more persistent than the pop-up shops and experiences that have proliferated over the past decade, executives say.

Wall Street Journal

Third places are seen as an opportunity for brands to garner attention.

Marketing executives today are also under more pressure to generate sales through their activities, rather than more nebulous metrics such as brand awareness and approval, than they were when Cadillac and others opened third places before the pandemic, Miraflor said.

Wall Street Journal

Quantitative trading from home

Quantitative algorithm trading is now within reach for non-quant investors.

Products like,, and are empowering retail traders to build, test, and run their own trading algorithms.

Composer went live in 2021 and enables users to construct strategies from a list of more than 11,000 stocks and ETFs. Its user base nearly quintupled in 2023 and now numbers in the tens of thousands, said Benjamin Rollert, the company’s co-founder and chief executive officer.

Wall Street Journal

Maneos, 44 years old, uses online-trading platform to build, test and bet on quantitative trading algorithms that buy and sell stocks and exchange-traded funds out of his home office in Boca Raton, Fla. One algorithm, for example, holds a triple-leveraged exchange-traded fund tracking the Nasdaq-100 index if the S&P 500 index has recently trended higher—and Treasury bills otherwise.

Wall Street Journal

Personal data as input for AI powered recommendations

This is an interesting use case for generative AI.

What if you used your reviews from Yelp, Goodreads, or Letterboxd to get personalized recommendations for restaurants, books, or movies?

It is easy to feel overwhelmed by the sheer volume of books, TV shows, movies, games, albums and podcasts out there. Reviews can tell you what is generally good, or not so good, but they can’t tell you what fits your specific tastes at this exact moment. Streaming services have a better handle on what you like—but they will only recommend stuff in their own catalogs.

Wall Street Journal

To get even fancier: Maintain a running list of titles and ratings on sites like IMDb or JustWatch (for movies and TV), Goodreads (for books) or your favorite videogaming platform. You can then paste numerous examples of what you want into the AI (“Here are some of the books I’ve given five stars”). Netflix and Goodreads even let you export your viewing or reading history as a file that you can upload to your GPT assistant.

Wall Street Journal

PWAs: Progressive Web Applications

David Heinemeier Hansson, cofounder of 37signals, recently wrote a post on how the time for PWAs may have arrived.

Basically it’s the idea that instead of multiple, distinct native applications, you can package one web application that feels (almost) native across all platforms. One code base, one deployment path. Massively simplified development.

Native mobile apps are optional for B2B startups in 2024

Previously a company could not get away with having a web only app. However the tide seems to have shifted, partly due to the European Union passing the Digital Markets Act in 2022, and the impact that is having on Apple’s App Store.

However we got there, here we are. The hallowed promise of web applications being viable and competitive alternatives to native efforts is finally here. Fans of the open web have been clamoring for this day for so long that it’s perhaps easy to miss that it’s finally here. But it is.

Native mobile apps are optional for B2B startups in 2024

This means that I’m now going to recommend that every B2B software startup begins by targeting the web first. Unless you’re swimming in a red ocean of competitors who are winning the market by stellar native apps, you can and should focus on getting to product-market fit on the web first. Then, once you’re minting money, and can afford the large teams required for multiple native applications, you can always explore the mixed blessings of App Store distribution.

Native mobile apps are optional for B2B startups in 2024