I'm not reading the details of this particular experiment, because I'm wasting enough time talking about it as it is. ;) But you have to understand the nature

I'm not reading the details of this particular experiment, because I'm wasting enough time talking about it as it is. ;) But you have to understand the nature of real prototype devices. Take it from someone who has built many semiconductor lasers that worked so badly that it took hundreds of thousands of dollars worth of specialized equipment to learn that they worked at all: The first transistor probably didn't work very well. The first laser didn't produce a lot of power or have a very long lifetime. Famously, the first several thousand Edison light bulbs didn't work well enough to be viable products. The difference between success and failure in an early-stage experiment is often small, and often statistical, and is often so ephemeral that you yourself have crises of faith in the results. And that's for the kind of phenomenon that eventually turns out to be tremendously important! And this is the problem. On the frontier, the difference between a promising sign and a statistical fluke is small. It's a territory ripe for self-delusion and fraud. You ask me to run a steam engine. Well, okay, suppose I can do that for five minutes. But then a piece of duct tape comes unstuck. You say: Replace that duct tape with something more durable. I reply, okay, sure, can you lend me the necessary $1500? (I was trying to save money with the duct tape. Half the parts are duct tape. Prototype device, remember?) You ask me to run the test longer. Well, okay, I can run it until my special catalyst runs out. You say: Get more catalyst. I reply, okay, sure, can you lend me the necessary $10,000? (The catalyst is expensive in small quantities, you see, but in the future we'll have economies of scale.) You ask me to open up my black box and submit every part to detailed analysis. We'll hire an independent laboratory and they'll handle all of it. I reply, okay, sure, can you pony up the necessary $10 million? And then the independent laboratory finds a suspicious anomaly. Oh dear, I say, I didn't realize that was happening. Maybe I made a mistake. I'll have to go back to the drawing board. Can I have a $1m grant to go back to the drawing board?