Sign In

Christian von Uffel

UX Researcher and Self-Taught Full-Stack Developer. Hosts monthly AI meetups in Austin Texas. Builds fun and useful software products.

Christian von Uffel is a seasoned professional with a diverse background in UX research, front-end web development, and digital marketing.

With his experience leading UX research at Celsius Network, a prominent crypto finance platform, Christian has honed his skills in conducting user interviews and usability tests to enhance user experiences.

Christian's educational journey includes studies in UX design, investment banking, finance, digital marketing, business technology, and more at reputable institutions like General Assembly, New York University, and Hofstra University.

Throughout his career, Christian has held various roles in organizations such as Self-employed, Perfecting Product, Celsius, Acceleprise, and GrowthX Academy, where he contributed his expertise in AI product engineering, UX design and research, growth marketing, and more.

Christian's hands-on experience in designing, building, and launching websites and applications, alongside his strong academic background, has equipped him with a unique skill set to excel in the ever-evolving tech and finance industries.

Highlights

Apr 15 · linkedin

Tech news outlets are coming out with critical reviews of humane, a new AI wearable device that lets users interact with LLMs to answer simple questions and perform simple tasks.

Unfortunately, most of these reviews are negative and highlight how the device fails to deliver on its promises.

The key reasons why humane seems to fail are two fold. First the LLM interaction doesn’t work reliably and is slow. Second, the device over promised what it could do.

Both these problems are innately solvable. LLMs have gotten tremendously faster as cloud inference offerings have grown and have become radically cheaper. This makes Humane’s LLM choice somewhat aggravating.

It’s highly likely that Humane chose to use a sub-par LLM for critical user interactions. But even if Humane fine tuned a mid tier model to improve its performance, choosing a cheaper model to save costs is a disastrously naive mistake. There’s simply a very critical threshold of reliability that needs to be met in order to deliver a good user experience. Below that threshold AI wearables should not ship.

AI wearables are getting a lot of attention for good reason. People want new ways to solve knowledge problems without distracting themselves with screens and getting eye strain. Humane could’ve hit it out of the park with their new product. Let’s hope the next AI wearable does a better job.

Christian von Uffel
Add to my network

Location

Austin, Texas, United States