I burned by 220,000 AI tokens simply because I might

I’ve been testing the $7,000 Asus ProArt P16 RTX Spark by throwing an unusually giant real-world analysis challenge at it. Reasonably than asking a chatbot questions, I’m utilizing OpenAI’s large 120-billion-parameter gpt-oss mannequin to chew by a whole lot of potential information factors, consider them in numerous methods, problem its personal conclusions, and spit out huge piles of study.

One early run examined 403 candidates and generated 220,202 tokens in just below 13 minutes. That’s a silly quantity of AI.

And since the mannequin was operating completely on my laptop computer, I didn’t care.

The meter adjustments all the pieces

These tokens weren’t free. A $7,000 laptop computer positive as hell isn’t free. Electrical energy isn’t free. My time isn’t free (sorry not sorry boss). For those who’re shopping for {hardware} particularly to keep away from cloud AI payments, you’ll want to account for all of that earlier than declaring victory.

However as soon as the laptop computer was sitting on my desk, producing one other 200,000 tokens didn’t produce one other API cost.

That adjustments how you utilize it.