[In order not to become a poison]

[In order not to become a poison]

The report that OpenAI, a hot company these days, is likely to lose $5 billion this year is a hot topic. The story is based on internal accounting information, and it is highly reliable.

OpenAI is expected to generate about $5 billion in sales this year, but it is expected to suffer a lot of losses due to the much higher cost, and the $10 billion investment last year may be exhausted by the end of the year at the earliest.

The issue with generative artificial intelligence right now is the lack of a definite "wow moment." It's fascinating and fun to try ChatGPT or other LLM services, but it's still a little lacking. It's difficult to replace call centers with artificial intelligence because 1% is anxious, and it's not easy to replace reporting or programming. It's true that it increases a lot of productivity, but it's difficult to create S/W coding for real products by relying on artificial intelligence. You still have to look in and finish it.

It is questionable whether generative artificial intelligence will spread as quickly as expected due to hidden costs such as changing the entire work process, maintaining quality, and concerns about accidents, even though it can save money if used well in other tasks other than programming. In fact, this is a problem that has been seen for years in the field of autonomous driving. The completeness of technology and the creation of new value are important, but it seems that more time is still needed.

Apple's Siri is the single platform used by the most people in the world and has the lowest artificial intelligence performance at the same time. It comes out even though I didn't call it, or I don't understand it, or I give a lot of wrong answers. If you make Siri that can be used with generative artificial intelligence technology, the situation can be greatly improved. You don't expect real-time interpretation and translation, and just listening to what I am saying in real-time can create great demand.

It's been more than five years since Tesla came up with the concept of FSD, but it hasn't yet made a clear gap with other electric car makers by self-driving and is still struggling to fill the last 0.1%. When the iPhone's Siri gets really smart, the next step will be humanoids that understand and perceive space, followed by self-driving, and I wonder how much money and electricity I will have to spend until then.