HN user

MrToadMan

125 karma
Posts0
Comments21
View on HN
No posts found.

Much like his earlier work Westworld which was also scarily prescient for modern times.

These are highly complicated pieces of equipment. Almost as complicated as living organisms. In some cases, they have been designed by other computers. We don't know exactly how they work.

The competence threshold can vary according to the model's complexity and cost to run, but the broader point remains that having a human operator of greater competence that can evaluate the output and understand potential harms that might result seems to be a valid one. Depending on domain, it can be irresponsible to deploy systems you don't fully understand that could harm others. Responsibly overseeing 37k lines of code a day sounds exhausting.

Leanstral 1.5 22 days ago

If only they had access to a world class translation system, they could auto translate between languages effortlessly :)

I guess my point was that I don't see many algo changes in the commit history, which is a shame if this has been lost; library/* files are largely unchanged from the initial commits. But each time the agent runs, it has access to the best solutions found so far and can start from there, often using randomisation, which the agent claims helps it escape local minima e.g. 'simulated annealing as a universal improver'. It would be nice to see how its learnt knowledge performs when applied to unseen problems in a restricted timeframe.

Not as many changes to the files under library as I expected to see. Most changes seemed to be under a single ‘add stuff’ commit. If some of the solvers are randomised, then repeatedly running and recording best solution found will continually improve over time and give the illusion of the agent making algorithmic advancements, won’t it?

What if while you were eating a cheeseburger, McDonald’s was magically replenishing that burger so that no matter how much you kept eating there was still some left. Moreover you had little control over the ratios of fat and sugar used to replenish it and they earned more the longer you spent eating it. Would you consider them harming you if they were prioritising stuffing it with ingredients that maximised the amount you ate and ignored sensible limits on sugar and fat?

Or if they took my deposit and commingled it with company funds and bought some illiquid luxury real estate in the Bahamas for their staff to live in totalling $240m. If they’re upfront about that, probably nobody deposits with them.

Is this as impressive as it initially seems though? A Bing search for the text shows up some Web results for Dvorak to QWERTY conversion, I think because the word ‘t.fxrape’ (keyboard) hits. So there’s a lot of good luck happening there.

From what I've read, about 1/3 of all asylum seekers over the last 7 years arrived via small boat crossings.

Looking forward though, about 90% of those arriving in small boat crossings are currently going on to seek asylum and the average annual cost of supporting an asylum seeker during their claim has risen to an estimated £41k, so for ~30k arrivals this year, the financial cost of not processing these claims promptly could increase that overall annual bill further still.

Also, in the first year of processing, costs may be drawn from the overseas aid budget (which was recently shrunk). This results in possibly 1/5 of the overseas aid budget being used for costs associated with processing asylum claims, which perhaps doesn't match most people's expectations as to what overseas aid should be used for.

I think that's why even though the number of people involved in these crossings is small compared to net migration, it has a big financial impact.

Depends on if you are looking at this in terms of numbers of people or cost. The Home Office annual spend on processing asylum seekers has ballooned from just under £1 billion to near £5 billion in the space of 5 years, which is 1/3 of the estimated £14 billion raised from the unpopular National Insurance increase.