Not as scary as the headline makes it sound, basically someone's made a better voip system. Article.
Still there are some impressive bits: 'For example, Voss says the system can use its ability to track the flow and sense of the conversation to work out who a pronoun - such as she or you - is referring to.'
And the article could be revealing how AGI might be priced in the future (based on how this is priced now: 'What's more, this brain is for sale with a tag of $30,000. It can also be rented for about 20 cents a minute.'
30k is a key price point. If you can buy AGI for 30k it will have a catastrophic and immediate impact on employment numbers. Why? Because a company would be able to justify spending on a system that will pay for itself in savings in just a single year (and thats assuming it won't also be eliminating jobs that pay over 30k a year- which it will), they might not be able to justify spending on a system with a longer pay back period, as this would too greatly impact their quarterly earnings.
Hopefully this price point doesn't reflect the cost of AGI systems in the future, and, hopefully, any incremental advance in AGI will come with an exponential increase to the price tag... but I doubt it. As always, I'm not worried about the AI or the AGI, just how smoothly its integrated into our economy.
Also of interest is the potential for early competition between humans and AI to push humans away from hourly wages and towards pay based on the amount of time spent on a given task. In other words, a lot of what an employee is paid today could be thought of as paying for their presences or paying for them to be 'on call' on site. If leased AI is able to do the same thing but the company is only charged with the amount of time the AI is actually working, will this force humans to accept similar terms?
Subscribe to:
Post Comments (Atom)
No comments:
Post a Comment