Hacker Newsnew | past | comments | ask | show | jobs | submit | elzbardico's commentslogin

AMD and Intel were focused on single core performance, because personal desktop computing was the bigger business until around mid to late 2000s.

Single core performance is really important for client computing.


They were absolutely interested in the server market as well.

And not everything in reality maps nicely to hypermedia conventions. The problem with REST is trying to shoehorn a lot of problems in a set of abstractions that were initially created for documents.

MCP servers allow controlled access to stuff that I don't want my agent to be able to code and execute at will.

I wrote a lot of small MCP servers in my Job that provide LIMITED access to stuff like AWS infrastructure, databases, SAP integration, Salesforce. By having coded them myself, I decide what access I give, what I deny, what I obfuscate, what I anonymize.

I am not trusting tens or hundreds of millions of liability to an LLM. LLMs have no operating world model, have no intent, they can't understand cause and effect, even if they can generate text that describes cause and effect.

I am a professional, I am responsible for the tools I use and run.


1) Raw inference speed matters more than incremental accuracy gains for dev UX—agree or disagree?

Yeah, I love reviewing and debugging thousands of lines of buggy and dirty AI generated code. Who cannot love it?


key word incremental - for fast apply to be useful it should be so fast and accurate that most people don't realize there's a model there at all

I work with AI, I consider generative AI an incredible tool in our arsenal of computing things.

But, in my opinion, the public expectations in my opinion are clearly exaggerated and sometimes even dangerous as we ran the risk of throwing the baby with the bathwater when some ideas/marketing/vc people ideas become not realizable in the concrete world.

Why, having this outlook, I should be banned of using the very useful word/concept of "hype"?


Your post doesn't contain a single prediction of a problem that will occur, dangerous or otherwise, just some vague reference to "the baby might get thrown out with the bathwater". This is exactly what I'm talking about, you just talk around the issue without naming anything specific, because you don't have anything. If you did, you'd state it.

Meanwhile the AI companies continue to produce new SotA models yearly, sometimes quarterly, meaning the evidence that you're just completely wrong never stops increasing.


> [...] when some ideas/marketing/vc people ideas become not realizable in the concrete world.

This is a single prediction of a problem that will occur. The tools not living up to the hype leads to disappointment, and people are likely to entirely abandon it because they got burned (throw the baby out with the bath water), even though the tools are still useful if you ignore the hype.


Really, the only reason the Kurds are bombed are themselves. They do bad decision after bad decision constantly.

Man. Huawei is fucking massive, they do far, far more things than just 5G base stations (a giant business in itself) and cell phones. They build even electric cars.

If they were trained on those benchmarks they would score 100%

They show how bad they are, they cannot score 100% on benchmarks they were trained on.

Jesus. What a steady consumption of American neocon propaganda can do to a human brain! It's so sad!

Do you even know what that word means? Do you think that Taiwan is gonna be just fine if the US packs up and leaves tomorrow? That things will work out great for the people living there?

You can call it whatever you want. People who have fled shitty regimes have a much better sense for propaganda than you do, evidently.


> Do you think that Taiwan is gonna be just fine if the US packs up and leaves tomorrow?

I do. The World will be just fine as the American empire fades and the US becomes just another country. Even for the American people, for the average persons, it will be an improvement.


Don't matter. You still have the debt. If you lose your job in a depression, odds are no much other people will be in the market to buy your home.

The amount of equity you have in your home matters. Most recessions do not take 20% off home values, so people with conventional loans are pretty safe. Even the Great Depression just cut valuations about 35%.

If you lose your job in a depression there will be plenty of people willing to buy assets at a discount. If you have equity in your home then your position will be net positive. About half of all mortgages have an outstanding balance less than 50% of the home's value.


That kinda works in the abstract, but it's pretty risky for any individual house. If your house was somehow a representative slice of all US housing, sure maybe it won't drop 35%.

But if you owned a house in Detroit from 2003-2010 it might have dropped 70-80%. Or, more on point for many on HN, if you own a house in the Bay Area worth $2-$3M with 25% equity, and the tech job market collapses, then you might get completely wiped out.


I think debt is good, actually, for most middle class households, and they're actively trying to increase their debt because that results in greater cash-flow, more savings, and more security in terms of retirement. That's why buying a home is Goal #1 for most Americans.

Home mortgage debt is mostly "good" debt for middle class households. Pretty much any other type of debt is a net negative in financial terms. There are less tangible benefits to owning a newer, more reliable car for example, so it can be a bit murky. General consumption debt is a bad bet for pretty much anyone.

Consider applying for YC's Fall 2025 batch! Applications are open till Aug 4

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

Search: