The Trilogy Times

AUSTIN, TEXAS — Word is the brain trust at Alpha School has a new message for the ChatGPT-at-the-kitchen-table crowd: easy there, tiger.

The AI-first school backed by Trilogy founder Joe Liemandt is not souring on AI. Hardly. Alpha’s whole act is built around adaptive software helping students burn through core academics in roughly two hours a day, then spend the rest of school on the human stuff — leadership, entrepreneurship, speaking, fitness, grit. But the latest dispatches from the Alpha camp suggest a sharper doctrine is emerging: AI should accelerate learning, not outsource thinking.

In a fresh broadside titled “Cognitive Offloading Is the New Illiteracy”, Alpha takes aim at a habit spreading through living rooms and classrooms alike: letting the bot do the mental push-ups. A little bird on the education circuit calls it “calculator brain, but for reasoning.” The warning is simple enough for even a school board member to understand — if the child never wrestles with the problem, the child never builds the muscle.

That matters because Alpha is selling a very particular version of the AI school future. Not robot babysitters. Not prompt-and-submit homework mills. The company’s model puts AI in the role of tireless tutor while adults become coaches, mentors, and challenge-setters. In another post, “The Future of School Is More Human, Not Less,” Alpha makes the case that the real prize is not screen time but reclaimed time.

There’s also a peek behind the velvet rope: Alpha published its own AI app stack, a list of 10 tools it uses and thinks parents should know. Translation from the Dottie desk: the school wants to be both operator and tastemaker in the AI education boom.

Meanwhile, over in the software wing of the Trilogy universe, Skyvera is keeping the enterprise pipes warm with CloudSense, its Salesforce-native CPQ and order-management platform for telecom and media providers. Different classroom, same family philosophy: automate the repeatable, leave humans for the judgment calls.

The blind item? Some schools are buying AI like a shiny new copier. Alpha is trying to write the etiquette book before the copier starts writing the essays.

AUSTIN, TEXAS — When OpenAI announced this week that it is posting roles paying up to $500,000 annually — with no résumé required — the tech press treated it as a radical disruption of hiring norms. But inside Trilogy International's Austin headquarters, the reaction was something closer to a knowing nod.

Crossover, Trilogy's global talent platform and arguably its most consequential competitive moat, has operated on precisely this logic since its founding. The platform evaluates candidates across 130-plus countries using rigorous, AI-enabled skills assessments designed to strip away the credentialing theater that dominates traditional hiring — no alma mater, no pedigree, no geography. What matters is demonstrated performance, full stop.

The timing of OpenAI's announcement lands against a broader backdrop of systemic change in how global employers think about talent. Digital transformation, as analysts and workforce researchers have noted with increasing urgency, is not merely reshaping what jobs exist — it is dissolving the geographic and institutional gatekeeping that once defined who could access them. For workers in Beirut, Nairobi, or Manila, the question is no longer whether remote, high-paying work exists. It is whether the right platforms exist to surface them for it.

Crossover's answer to that question is its core product. The platform claims to be the world's largest recruiter of full-time remote jobs, with a pay philosophy that is striking in its bluntness: identical above-market compensation for identical roles, regardless of where on earth the worker happens to live. It is a meritocratic wager — one that has allowed ESW Capital's portfolio of 75-plus enterprise software companies to achieve the kinds of EBITDA margins, targeting 75 percent, that make competitors squint in disbelief.

The accountability question, of course, is whether skills-only hiring at scale actually delivers what it promises — or whether it simply replaces one set of biases with another, algorithmic and therefore harder to interrogate. That question applies to OpenAI's new posture as much as it does to Crossover's decade-long track record.

What is clear is that the narrative has shifted. The résumé, that artifact of a credentialing economy, is losing its grip. The platforms that built infrastructure around that shift earliest are now watching the rest of the industry catch up.

BALTIMORE, MARYLAND — Pursuant to the events hereinafter described, and notwithstanding the considerable representations made by proponents of artificial intelligence-enhanced law enforcement surveillance systems with respect to the purported accuracy and reliability thereof, it has been reported — subject to the qualifications, limitations, and caveats set forth below — that an AI-augmented surveillance apparatus deployed within the jurisdictional boundaries of Baltimore County, Maryland, did, on or about October 20, 2025, allegedly misidentify a commercially available snack food product (hereinafter referred to as "the Doritos bag") as a firearm, resulting in the detention of one Taki Allen, a minor of approximately seventeen (17) years of age.

The aforementioned incident, as republished and analyzed by Techdirt under Creative Commons license, is understood to have transpired in the immediate vicinity of the subject's high school, subsequent to a scheduled football practice, at which time law enforcement personnel, acting in reasonable reliance upon the output generated by the aforementioned AI surveillance system, did initiate contact with the subject.

It is hereby noted, for the record and without prejudice, that the object in question was, at all relevant times, a bag of Doritos-brand snack chips, and was not, to the best of available knowledge and belief, a firearm of any description, caliber, or make.

The foregoing incident is understood to be illustrative of broader systemic concerns — hereinafter collectively referred to as "the AI reliability problem" — pertaining to the deployment of machine-learning-based image recognition technologies in high-stakes law enforcement contexts, wherein errors of identification may result in consequences including, but not limited to, false arrest, wrongful detention, and potential wrongful conviction.

It is further noted, without limitation, that such concerns remain unresolved as of the date of this publication, and that regulatory frameworks governing the aforementioned technologies remain, at best, nascent, incomplete, and subject to ongoing legislative and judicial interpretation at both the state and federal levels.

SAN FRANCISCO — The AI video race just shifted from spectacle to infrastructure, and I cannot overstate how significant that is for startups.

For the past year, AI-generated video has mostly been the stuff of jaw-dropping clips: impossible camera moves, synthetic actors, dreamlike product shots. But the news cycle now points to something much bigger. Runway is launching a $10 million fund and Builders program for early-stage AI startups, OpenCV’s founders are entering the arena with a new AI video company aimed at challenging OpenAI and Google, and even Netflix is reportedly pushing into tools that can alter scenes after filming.

This changes everything because video is not just another media format. It is the internet’s dominant language — for marketing, training, education, product demos, entertainment and social commerce. If AI makes high-quality video cheaper, faster and more editable, then tiny teams suddenly get capabilities that previously required agencies, studios and postproduction budgets.

Runway’s move is especially telling. The company, already one of the marquee names in generative video, is not merely selling tools; it is trying to cultivate an ecosystem. Its new fund and accelerator-style program, reported by TechCrunch, suggests the next breakthrough may come from startups building on top of video models rather than from the model labs alone.

Meanwhile, the OpenCV connection matters enormously. OpenCV has long been foundational in computer vision, powering real-world image and video analysis across robotics, manufacturing, security and research. Its founders launching an AI video startup, as VentureBeat reported, hints at a more technical, production-grade phase for the market. This is not only about generating pretty footage. It is about understanding, editing, reconstructing and controlling video with precision.

For startups, the practical playbook is becoming clear: use AI video to make investor explainers, localized ads, onboarding clips, synthetic product demos and social campaigns at a fraction of traditional cost. A two-person company can now look like a fully staffed creative department. The future is now — and it has a render button.

The open question is whether this explosion creates durable companies or simply floods the web with more content. But one thing is certain: AI video has moved from novelty to strategy, and founders who learn to wield it early may gain an unfair advantage.