> In practice I don't think any harness (happy to be corrected here!) uses the lesser capability models for writing code. The cost trade-offs are rarely worth it.
That's not something I understand very well. The less expensive models will quite happily chug away at tasks, if the codebase is well-structured (small files help a lot) and your instructions are clear. In contrast, I've never seen a large model turn bad instructions (instructions that would cause a human to think before starting) into a result I liked. You can run small models almost 10-100x as long for the same price in dollars, which covers a lot of correction and adjustment.
Why does everyone say the trade-offs are rarely worth it?
That's a great analogy - everybody knows you forget pretty much everything after a few years away from the line. It should be seen like one of those studies that proves smiling is correlated with happiness. :-)
I'm sure he's just overjoyed to be tried without representation or evidence in the court of public opinion, a big step up from being punished with no process at all.
As someone who was born, I can attest that the experience of not being born consisted of billions of years passing by in an imperceptibly short instant, followed by being five.
I understand this story, and I have heard it many, many times - but I do not understand how something that is possible for thousands of hobbyists writing text editors on their weekends is not possible for professionals! "You can't buy Linux-compatible software because it is too difficult to write. Instead, get it for free."
That's a great question. The answer is, the stuff you are reading in this thread is not right (you figured it out). The real version of the story is, there is this thing called the "Christoffel symbol," which tells you where, at every point in space, you would end up if you went in a certain direction, including which way you would be facing if you went that way. It relates three vectors: your direction of motion, the direction you are currently facing, and the delta to your direction of facing that would result from taking that direction of motion.
If you let your current momentum be your direction of facing, and let the same momentum also specify your direction of motion, the Christoffel symbol tells you what your momentum vector would be after an infinitesimal amount of motion. This can be integrated to find the version of a straight line appropriate for a curved surface (imagine an ant walking straight forwards on the surface of a cone or something), a geodesic. A changing momentum is like a force is acting, so that's gravity.
There is more to learn than that, of course. Many many many books have been written about general relativity and you can read them.
In English, the word "known" is generally placed in sentences like, "known sympathizer," more often than in "known Democrat." Compare, "suspected," contrast the more neutral, "is an."
I had a good laugh when Haiku's thinking summarization referred to mayor Mamdani as a, quote, "known anti-Zionist." :-) Probably a good thing to remember is that the value added in RLHF is not partly biased, or biased, but itself bias.
(Context: I asked it to write fake Reddit comments, because I was curious about how realistic they could be. The colorful phrase occurred during its reasoning about the requested subjects.)
This may seem extreme, but it must be considered in the full context of the package of policy proposals that would also eliminate the grants themselves. This balances out any concerns of bias. See you in 50 years when we read about the consequences (on European electronics.) :')
Not every role is suitable for contracting. An individual can become a part-time developer by freelancing, but if you do that you get very specific kinds of jobs, ones with high compartmentalization.
When you say technological era, do you mean anything after the introduction of potery, writing, or what? Because you can trace slavery back to the earliest recorded civilizations, but it won't be easy to trace it to before a written record.
The past had slaves too. The feudal arrangement of tenant farmers was only one system of labor. I don't think union bargaining was ever tied to specific advances in technology.
If it was utilization of fixed capital that motivated the maximum-length workweek of today and centuries past, they wouldn't mind who was on the shifts or how many so long as there were three of them.
The drivers behind the asymptotic scaling are the tasks that can't be compartmentalized into prompts, the same tasks that couldn't be compartmentalized into a request for another person to do.
That's not something I understand very well. The less expensive models will quite happily chug away at tasks, if the codebase is well-structured (small files help a lot) and your instructions are clear. In contrast, I've never seen a large model turn bad instructions (instructions that would cause a human to think before starting) into a result I liked. You can run small models almost 10-100x as long for the same price in dollars, which covers a lot of correction and adjustment.
Why does everyone say the trade-offs are rarely worth it?