Software Engineering Daily

SED News: Apple’s AI Problem, The Real Business Model of AI, and Token Cost Reckoning

51 min

This episode of Software Engineering Daily explores Apple's AI struggles, the true business model of AI companies, and the impending token cost reckoning. The hosts discuss Apple's fake Siri demo and its hardware-focused culture, while analyzing how AI companies like Anthropic and OpenAI are finding profitability through enterprise subscriptions rather than consumer plans. The episode also covers Google's AI search pivot, Remote's $300M ARR milestone with flat headcount, and the addictive nature of AI coding tools.

Summarized by Podsumo

🎧 Listen 🎙️ Ask about this episode

✨ Key Takeaways

1

Apple faces a class-action lawsuit over a fake Siri AI demo, highlighting a cultural gap between its hardware-optimized business model and the innovation-driven AI landscape, with a 250 million dollar lawsuit filed against false advertising.
2

AI companies like Anthropic and OpenAI have found product market fit through enterprise subscriptions, where a $100/month plan can consume $2,000 worth of tokens, suggesting consumer pricing is subsidized by enterprise spending.
3

Remote reached $300 million ARR with flat headcount, indicating AI-driven efficiency, but there are concerns this may simply mean employees are working longer hours rather than genuinely becoming more productive.
4

Google’s default AI mode in search has led to a 28% increase in DuckDuckGo traffic, as users seek less ad-heavy alternatives, signaling a potential shift in the search landscape.
5

The 'Doom on a travel router' tradition persists, with a new port on a GL.iNet Slate 7 Pro, and someone has created 'Doom Bench' to test if your stack can run Doom.

💬 Notable Quotes

"Apple warned that a truly capable AI assistant could make manually downloading apps obsolete, since services (like the App Store) are a huge money driver."
"I think it's interesting because Apple is such a very successful company... but it's kind of hard to see what their next act is."
"This isn't an AI failure story. It's just that a budget set in 2025 failed to predict how indispensable these tools would become."