Forgot your password?
typodupeerror

Comment Re:Making Money Fixing AI Code (Score 1) 74

No, you decided to throw a reference that required interpretation, hoping the reader was aware of Rummy and his views on information. Fortunately I was, and you can't test a negative case you don't know about, that's why it's an unknown unknown. If I were a known unknown, there are lots of tests.

Comment Re:Making Money Fixing AI Code (Score 1) 74

Unknown unknowns are part of the bet unfortunately - DD is done reasonably rapidly, things do get missed naturally - usually it's a synergy category error in retrospective analysis. But given it's a blend of people, technology, culture in org - and these can't be neatly tied with a bow like coding. You can go and read your AI's code and correct it, an acquisition or disposal doesn't have such surety - judgement is needed. And that judgement sometimes does fail.

Comment Re:a380 concorde (Score 2) 61

Depends on the route - there are some that make it a highly desirable airframe. But they are definitely far fewer than they need to be. It's not correct to contrast with Concorde - that was always a prestige vehicle, it never made any meaningful dent in passenger-carry terms. A380 is a post-Panamax vessel in a Panamax world.

Comment I suspect its got another 24 months (Score 4, Interesting) 64

24 months to deliver an ROI - or the wheels on the bus start coming off. Especially given the additional commodity expenses we now face. That doesn't mean AI will disappear - it's way too useful for that - but it will see a rationalisation. Railway Mania ultimately led to latter rationalisation and branch lines closed. So I suspect it will be for datacentres. Costs of use will go up, quality in narrower fields will improve, the race for generalised AI will cease. We rather got ourselves into a heavy lift race - AGI was no Moon to capstone the race.

Slashdot Top Deals

"Plastic gun. Ingenious. More coffee, please." -- The Phantom comics

Working...