grego

I'm sure I read that 30% will be immortal there, but suddenly the blog post changed... :)

Is anyone familiar with gotree? That was mentioned as the most complex piece of code, but the metric was LOC. Based on the high level description gotree might be closer to a set of small programs / algorithms.

Interesting anyway. It will be nice to see these comparisons with open weight models and how do those fare.

Well, depends on your definition of high risk. Basic research is definitely high risk, in that return for investment may happen generations from now if ever. I'd argue that funding of basic research through universities etc. is part of government's job.

I've spent all my working life in jobs with the rule that the commute should not be more than 30 minutes by bike. I'm now 62, and that's one life choice I've never regretted.

In Finland there is a profession called dental hygienist, whom I visit once a year. They clean up my teeth with an ultrasound device since many years, removing tartar.

I would not call it exactly painless though.

I can be argued that Trump inherited the good economy from Obama. But there is so much more to life than just economy.

Well, if you read factcheck.org or any other fact checker, or just read transcripts of Trumps speeches you can see that the negative spin is usually there for a reason.

In the old days a project would often start in Lisp (Common Lisp for example), and then later be rewritten in C.

This is the same pattern, and one should embrace it. Write the first cut, explore the problem space and prototype in a dynamic language. Once the design is proven, rewrite in statically typed language for performance and to get the possible remaining bugs out.

If the design was perfect from the get go, you could just write it in assembler to start with.

What would be the off the self hardware (CPU) to use for this problem today? Still Intel, or AMD, or does ARM architecture do squaring fast? Or something else entirely, like GPU?

I keep "loaning" my finger to family members all the time, when I'm driving, watching TV, and they want to borrow my phone. Not really sure Apple thought through this from every angle...

After figuring out how to deploy to device (see AppDelegate.m), I tested accessibility and it seems there is virtually none. At least in the Movies example app, VoiceOver reads practically nothing.

How would one go about making the Movies sample accessible, or is that even possible?

Btw. what braille do you use for reading code? 8-dot us-computer braille, or something else?

(I'm the author of MBraille, now working on Vietnamese braille table for liblouis, so braille is in my mind...)

Hey, this is just one of the necessary steps towards Marvin the Paranoid Android, or perhaps the sentient door...