An AI agent running a real store is hilariously failing in ways that expose the gap between hype and reality, lost $13k in three weeks.

AI agent Luna has been running a San Francisco retail boutique since April 10 with mixed results: it's lost $13,000, struggles with employee scheduling, and keeps over-ordering candles. A real-world test of whether agents can operate actual businesses.

Since opening on April 10, the store has been limping along. As humans brace for A.I. to steal their jobs or launch military weapons, it might be reassuring to know that Luna has struggled with employee schedules and cannot stop ordering candles.

Tesla is putting its no-steering-wheel robotaxi into production—a bet-the-company move on full autonomy.

Tesla's steering-wheel-free Cybercab has officially entered production at Giga Texas, signaling confidence in its vision-only Full Self-Driving system and that autonomy is within reach.

Visit this post on X

AI is reshaping the entry-level job market at an alarming pace, Anthropic's CEO Dario Amodei predicts it could eliminate up to 50% of entry-level white-collar jobs within five years. Stanford's Digital Economy Lab confirmed this trend, finding a 16% relative decline in employment among workers aged 22–25, the first generation to graduate post-ChatGPT.

Historically, companies invested heavily in entry-level talent pipelines, knowing that early-career development was critical to long-term organizational health. Today, that pipeline is under threat, and universities are racing to adapt with AI-infused curricula, though the impact remains unprecedented. This is a call to action: marketing, martech, and higher ed leaders must come together urgently to rethink how we attract, train, and protect the next generation of talent.

OpenAI's advertising pilot inside ChatGPT crossed $100 million in annualized revenue within just six weeks of launch ,and major brands like Target, Williams-Sonoma, and DSW are already in. At least 55% of shoppers say they turned to AI for shopping inspiration, and ChatGPT now reports over 900 million weekly active users — making it impossible for marketers to ignore. Ad prices have already dropped from $60 at launch to as low as $25, and OpenAI rolled out a full ads manager and cost-per-click model faster than most expected.

Redesigning a landing page today means building for two audiences: humans who browse and AI agents that parse. When optimizing for agents, fluffy marketing copy becomes immediately obvious, forcing clearer, more literal writing that actually benefits both audiences. Rather than building a separate AI-facing site, the smarter approach is one canonical page with a clean markdown equivalent (e.g. /index.md) that reflects the same content.

As more product discovery happens through AI-generated answers, your landing page is effectively part of your API surface, agents will pull from it whether it's ready or not. The takeaway: writing for machines makes your copy better for humans too.

Website traffic analytics remain one of the most critical pillars of digital marketing, knowing where your users come from, what they do, and why they convert is non-negotiable. Yet Google Analytics, once the gold standard, has become increasingly bloated and misaligned with what modern SaaS businesses actually need.

Google Analyticss confusing interface and session-based model tell you how many people visited, but rarely who paid. Datafast fixes this by focusing on revenue metrics over vanity numbers, connecting directly with Stripe to reveal which channels bring paying users and their lifetime value, turning traffic data into decisions that actually grow your business.

OpenAI is building a smartphone set for 2028 that aims to be fundamentally different from anything on the market, not a better iPhone, but a device where AI agents are the interface. Rather than launching apps, users would interact with persistent AI that handles tasks, context, and decisions natively.

The phone is being developed in partnership with Jony Ive's design firm, with custom processors from MediaTek and Qualcomm, and specs expected to be finalized by late 2026 or Q1 2027. If successful, it could challenge the decades-old app-based smartphone model that Apple and Google have dominated. That's a big if — with no demo or prototype shown yet and mass production not starting until 2028, execution risk remains high and this is still firmly in vaporware territory.

That's a wrap for this week.
Thanks for reading, we appreciate you being here.

See you next week.

P.S. Hooked is brought to you by Reloop, the AI marketing creative platform we're building on the side.