> Using Bloomberg’s numbers, ChatGPT does NOT appear to have a racial bias when it comes to judging software engineers’ resumes.2 The results appear to be more noise than signal.
Which in most contexts means the same as "does appear to not have a racial bias", but not in statistics. One of the reasons why communicating results in research accurately is incredably hard.
They also said "that there was, in fact, no racial bias", which is a bit stronger than "no evidence of racial bias". In a context where words like "significant" are overloaded, it makes sense to me to be extra careful with phrasing.
GC feature in wasm extends to having opaque references outside linear memory (which may reference objects controlled by host system's gc) avoiding need for js caller to handle resource being freed by wasm code with a pool or something keyed by handles (integers) passed to wasm
And it's so much easier to pay by phone! I can be through the payment process in 20 seconds at a supermarket if I pay with my phone, and need nothing on me than my phone. Meanwhile cash takes fumbling, counting either by a machine or by a person, calculation in my head what I need to provide, making sure I always have not too little and not too much on me, making sure the total amount of coins doesn't get out of hand, etc.
It removes the incentive for clients to use cash if a card option is available, makes them more used to paying by card, and hence decreases the competitiveness of merchants that don't offer card at all - until cards become so wide spread that many merchants don't even accept cash at all, like where I live.
A bit hostile towards merchants, but very nice for consumers imo.
I spend primarily on credit cards but I still think it’s important to preserve the ability to transact without the government knowing. Privacy is best preserved when data is not created at all, rather than when data is protected by law.
I don't understand enough about the subject to say, but to me it seemed like yes, other models have better metrics with equal model size i.t.o. number of neurons or asymptotic runtime, but the most important metric will always be accuracy/precision/etc for money spent... or in other words, if GPT requires 10x number of neurons to reach the same performance, but buying compute & memory for these neuros is cheaper, then GPT is a better means to an end.
Glider guns have a period of 30, so you could work them out on one tablet of graph paper if you wanted. Or just let a computer do it frame by frame and pay attention.
This isn't a rebuke at all btw, it's a suggestion, this is a fun and edifying way to spend some time. Don't know if your kids are old enough for it yet, or have the interest to play along, but you might enjoy it.
> Using Bloomberg’s numbers, ChatGPT does NOT appear to have a racial bias when it comes to judging software engineers’ resumes.2 The results appear to be more noise than signal.
Which in most contexts means the same as "does appear to not have a racial bias", but not in statistics. One of the reasons why communicating results in research accurately is incredably hard.
reply