I disagree. Designing, testing, and building hardware is extremely expensive, and has a really long range return on investment; the fact that they coupled that

I disagree. Designing, testing, and building hardware is extremely expensive, and has a really long range return on investment; the fact that they coupled that with AI, also a long term ROI, means that they really needed a great deal of money at having a chance to succeed. I could see Cozmo as play to bring in a little more capital until they could finish creating industrial grade robots, but possibly too little too late. I'm not saying you need $200M to design/build hardware like this, but in this case it seems they may have expanded too quickly and didn't capture enough market share.