Algorithms

Explanations of recommendations, ranking, personalization, and automated content filtering. We look at what kinds of data algorithms usually take into account and why results can differ from one user to another.

Comparison of Reddit thread structure with extractable passages versus blog post format showing fewer citable sections
Algorithms

Why AI Search Keeps Quoting Random Reddit Comments Instead of Your Blog

Ask a friend which washing machine to buy and you’ll get something messy and specific: what they own, what broke after eight months, what they’d do differently next time. Ask a glossy brochure and you get spec sheets and adjectives. For most of the web’s history, search engines behaved like brochures, ranking polished pages written to be found. Now the AI systems answering people’s questions increasingly behave like the friend group — and when they go looking for that kind of testimony, they overwhelmingly land on Reddit.

A security camera and airport checkpoint illustrating algorithmic ableism in AI surveillance
Algorithms

When “Normal” Is a Setting an Algorithm Chose For You

Imagine walking through an airport with a security camera that isn’t checking whether your face matches your passport, but silently rating whether your gait, your eye contact, and your speech pattern look “normal” enough to pass without a stop. You never see the score. You don’t know it exists until, maybe, someone pulls you aside. For most travelers this system is invisible because it agrees with them. For a person with Parkinson’s, autism, or a stutter, it may be quietly disagreeing with their body the entire time.

AI shopping assistant comparing tidy retail storefront data for an AI commerce rankings benchmark
Algorithms

The New Scoreboard That Doesn’t Care How Much You Sell

Imagine two shops on the same street. One has a beautifully organized storefront, clearly labeled shelves, and a staff member who can instantly tell any customer what’s in stock, in what sizes, and how to buy it. The other is a sprawling warehouse with excellent products buried under vague signage and no one at the counter who can answer a simple question. A human shopper willing to dig might still find what they need in the second store. An AI assistant, скорее нет — wait, write in English only.

AI authentication checking a luxury watch for genuine details, showing how an AI certificate can prove a watch is real but not predict future value
Algorithms

Why an AI Certificate Can Prove a Watch Is Real but Not That It’s a Good Bet

Imagine a locksmith who can tell you, with total confidence, that the key in your hand is genuine — cut on the right machine, stamped with the right code, indistinguishable from the original. Now ask that same locksmith whether the house it opens will be worth more in ten years. He’ll shrug. He’s an expert in keys, not real estate markets. That’s roughly the position luxury watch buyers are in right now, caught between two forces moving at once: tariff-driven price shocks reshaping what watches cost, and a wave of AI-powered authentication tools promising to tell buyers what their watches really are. Both are useful. Neither answers the question people actually want answered: will this thing hold its value?

A shopper comparing products while an AI decision layer filters which brands appear first
Algorithms

The Invisible Shopper: How AI Decides Which Brands Even Make the List

Picture walking into a store where the assistant has already skimmed every product on the shelf, cross-checked the reviews, compared prices at three competitors, and formed an opinion about which brand is trustworthy — before you’ve said a word. You just ask, “what’s the best option?” and get an answer that feels neutral, even authoritative. That is roughly the position brands find themselves in now that AI search tools and shopping assistants sit between a customer’s question and the open web. The assistant has already done the browsing. Your job, if you run a website or a brand, is to have been legible to it in the first place.

A laptop showing Google search results with competing ads beside a branded query, illustrating branded search competitor ads
Algorithms

When Your Brand Name Becomes Someone Else’s Ad Opportunity

Imagine walking into a store because you saw your favorite brand’s sign outside, only to have a salesperson from a rival company greet you at the door and steer you toward a different shelf — without ever saying a bad word about the brand you came for. That’s roughly what happens when someone types your company’s name into Google and a competitor’s ad appears next to yours, sometimes without your brand name ever printed in the ad itself. It isn’t always a rule-breaking move. Often it’s just how the search auction, and the page it produces, actually works.

Scroll to Top