HN2new | past | comments | ask | show | jobs | submit | Daishiman's commentslogin

Well yeah. I went to a world-class research university in Latam. The brightest kids were making using of internships to Google, Microsoft and Meta a couple years before graduating and by then they already had a career path laid out for them in the Bay Area.

I'm sure a bunch of companies took advantage of the H1-B program, but without a doubt it took most of the best talent too.


It's bifurcated. Big Tech used it to increase the labor pool local to their existing development centers. The consultancies used it to drive down unit costs.

You work on niches that have very specific requirements that you can only derive from having a good relationship with customers and so you attend to those needs faster than competitors who are out of the loop.

> you attend to those needs faster than competitors

I wonder if this type of hustling can be called moat building?


The question to ask is, better than which human?

The top 25% percentile of coders in my org definitely code better than most agents. The rest? I trust the output of an LLM to be far more consistent when adding/deleting features across service layers than a human that can create accidental typos. Same thing with bog-standard React components or Docker build scripts.


> I've watched teams go from deploying weekly to deploying 5x/day after adopting AI coding tools. Their velocity metrics looked incredible. Their incident rate also tripled. Not because the code was worse per se, but because they were changing more things faster than their observability and testing could keep up with.

But this an improvement! The features/incident rate improves as you have more incidents but fewer in relation to the increased velocity. This may or may not be a valid tradeoff depending on the impact of incidents.

At least in my org we have an understanding that the product side will have to change drastically to accommodate the different rates of code development.


Documentation can be fairly rote, straightforward and can have a uniform style that doesn't benefit from being opinionated.

Power exists whether you like it or not and when power gets away from decisionmaking you just generate a power vacuum.

Power needs to be placed in the hands of better decision-makers. That starts from getting money out of politics.


Paving the way for arbitrary detentions and concentration camps.


These are already arbitrary detentions, well unlawful at least (they're purposeful, just not for legal purposes).

The tens of thousands of detainees aren't being put in hotels... they're going to concentration camps; either in USA where they're forced to work (slavery you might term it, as many (most?) have not broken the law, nor been detained legally); or abroad where the regime's intention appears to be that they die.


Truly a massive disappointment.

So many intellectuals that appear to dedicate their lives to developing coherent and consistent models of how to think and how to act in this world, who nonetheless decided that conversing with a known child rapist is A-OK.

The only people I've seen who have come out ahead morally are Nassim Taleb and Norman Finkelstein.


People have personae. They can have good and bad sides. On the evening news parents or neighbors will often say “he seemed like a nice kid; I never saw him do anything bad.” About some kid (or older adult) now being involved in some crime.

Don’t make heroes out of people. Many have significant flaws, even unforgivable flaws.


Something I think we should teach ourselves to do is never believe we know someone until we see their character tested. It's very easy to present as having good character under typical circumstances, but much harder when you're pressured by stress, duress, lust, impulse, or other forces. A good measure of someone's character is how they respond to these things, not only how they function in ideal circumstances.

Of course, there are more common measures like how they treat people they don't necessarily need to respect, but even that can be conjured on demand according to situational needs by cunning people. Getting to know people and drawing conclusions about them should be a longer process than it tends to be.


Money and sycophancy also corrupt behavior. Another measure of character, is the response to being broke.

In any case Bill Cosby on TV was a good role model -in life, out of the public eye, he was a bad dude who drugged and SA'd victims. The public has little chance to "know" him. One might hear rumors but one needs factual info to make judgements.


> On the evening news parents or neighbors will often say “he seemed like a nice kid; I never saw him do anything bad.”

If you're ever interviewed by the news, you have a moral imperitive to say "They were always such a good kid. I would have never suspected this." regardless of if you know the kid, or you would have suspected it. A) who needs reporters in neighborhoods digging up dirt like this. B) if it were you, or your kids, you'd want people to say the same thing for you.


Well, when I hear "people have personae" I'd imagine something like "This person is an esteemed professor at MIT but he's also a regular in erotic fanfiction forum," not "he's friends with a child sex trafficker."


And Jim Simons also comes out of this looking good.


Were there comms between them or one referencing the other?


There was a comm where Peter attia asks Epstein if he knows Simons. Epstein replied, never met him.



With some experience you start getting the hand of how big a task you can hand off, and then you give the agent a way to test the changes, like using a browser and checking for a component, or adding unit tests for the backend.

Having a tight feedback loop for agents is critical for getting good output.


A size like that is going to be completely, absolutely obliterated by micrometeor collisions.

These people are all smoking crack.


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: