tomaszcarter

He has a serious cardiovascular problem that they are hiding. He cannot spend a long outside in the cold because it will become very obvious.

Or… it’s a bluff. Probably is to be fair: their psychology is kind of transparent at this point.

Sounds to me like he thinks they have exactly fifty senators on side. Which is shocking - implies one of McConnell collins murkowski or fetterman has signaled their capitulation.

A bit more plotting might have solved a lot of later problems there.

What’s actually happening: trump doesn’t give a toss about rfk. In fact would probably prefer him not to get through, so kennedy’s team is having to do the dirty work. Rfk was conned - nomination for secretary does not equal getting the post. He will not last even if he gets in.

you have to read everything he does through the prism of “I want to be the top trend on twitter”. For a week he’s had all the attention and suddenly this thing comes along from china. So he needs to insert himself in the story, somehow.

the advent of deepseek means that it has almost no chance of it going away. The main thing holding it back was the cost. If what this latest model claims is true we now have cheap AI at our fingertips. This also means headroom to increase the accuracy again at a reasonable price.

this is my concern! It feels like these conversations start from a place of distrust very often when really we are all in it together and finding the best uses for this new tool should be our aim

Well if you want to ask it that way I can only really explain in metaphor: you now have two people doing the same project, one real one artificial. They complement each other in terms of what they know and one of them can type really fast. Real coder has refined instincts, ai coder better memory.

I think people aren’t trying hard enough to find the appropriate middle ground. LLMs are neither the all conquering answer to every human dilemma (yet) nor a complete waste of time and money. Refusing to explore the possibilities is fine but you will miss out.

No. It looks like actual measurable output being at least two times higher, a measurable improvement in quality and the ability to work in languages I have never used before. You miss the point that AI output is way more human parseable and is replacing Google/stack overflow with a superior product.

It is possible that it depends on the model you are using.

It’s incredibly hard to gauge the total. I would estimate that I am 2-4 times more efficient with an ai coding partner than without. I am delegating: code reviews, syntax look ups, architectural suggestions, bug finding, writing tests

Absolutely not in my experience. I ask it a coding question. It answers, i am suspicuous or don’t understand, i ask it for the original docs and it provides them.

My experience is not this (coding, health, travel). You can treat it as a relatively senior person who will make mistakes. It is very dangerous to give it to people who don’t know what they are doing and ask them to use it to craft professional answers. Not so for professionals who want to save time

Seems like you haven’t really done much research here: chat will provide excellent links to original sources if you just ask it to.

Hmmm. The answers here are kind of depressing. At the very least, LLMs are much better replacements for Google without the sponsored links, that maintain some kind of context so you can follow up with a deeper question. They make mistakes (like Google does) but they get a lot right too.

He’s got some serious form of cvd or heart problem. The only reason to keep him out of something not that cold. watch for respiratory issues while speaking, swelling in lower legs (hard to spot) and discoloration in his face (visible through the orange?). Vance will be 48 this year or next.