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

> Benchmarks are toys, real world use is vastly different...Why should anyone waste time on poorer results? I'd rather pay my $200/mo because my time matters.

This kind of rhetoric is not helpful. If you want to make a point, then make one, but this adds nothing to the conversation. Maybe open source models don't work for you. They work very well for me.


Just go to a popular YouTube retirement channel and check the comments on a new video. For instance: https://www.youtube.com/watch?v=Awtflo0cU4o

Most of the creators delete those comments, but if you get in early, you'll see them.


Thanks! I think I found one in that link.

@DeniseJ.Wilson 6 hours ago I believe that a 'hard economy' is actually the best time to build if you have discipline and the right strategy. Most people spend their time complaining, but I believe in execution over excuses. I made a smart move in early 2026, and because I believe in staying the course, I just hit $1M. The momentum is still building, and by the end of this year, it’s going to be massive.

And the replies start talking about a Coach Julia out of no where. To search set up an apt.


If only Google had access to an AI service. That sounds like a snarky comment, but given how easy it would be for Google to use AI to zap those, I can only think there's something preventing them from using it.

> a retirement vehicle that doesn't let you withdraw until age 59.5.

If you're referring to US retirement accounts, that's not accurate. The early withdrawal penalty is 10% - the same as jumping from the 12% to 22% tax bracket when you're working.

If your company allows partial withdrawals starting in the year you turn 55, you can use the "rule of 55" to get your money out penalty-free January 1 the year you turn 55.

You can withdraw Roth contributions penalty-free at any age.

You can take SEPP withdrawals without a penalty.

You should have some cash and brokerage account money too. You could also own a rental house, sell your house and become a renter, etc. The 10% penalty is seldom going to stop someone from retiring.


> * The early withdrawal penalty is 10%*

Still cheaper than my current Long-term Capital Gains tax rate too.


Seems obvious to me that China would not want to give the AI market to US companies. You don't even need anything like an attempt to "sabotage the West". If I were them (the companies or the government) I'd be very hesitant to let US companies dominate this space. Especially companies that close to the current US administration.

Exactly, more large nations should be establishing or fostering their own labs. Outside of the Chinese and US companies there's really only Mistral.

I've always enjoyed using Clojure. Unfortunately, most of the things I do require interacting with the C world, so it has never been a real option as my primary language.

You might be interested in https://jank-lang.org/ - still early days but full C/C++ interop is the plan.

Coffi [1] built on java 22 (project panama) C FFI. Makes binding to C from Clojure a lot of fun.

- [1] https://github.com/IGJoshua/coffi


Have a look at Janet. It is not as powerful as Clojure but for smaller projects, prototyping, scripting etc., it is really enjoyable.

https://janet-lang.org/


I also interact a ton with C and C++, but it's easy today to have Claude write a Project Panama wrapper and then put a nice Clojure veneer on top of the Java.

Some feedback: It would be useful to explain what you do differently on your website.

Based on what's in the article, it wouldn't take much to move these files to SQLite or any other database in the future.

Edit: I just submitted a link to Joe Armstrong's Minimum Viable Programs article from 2014. If the response to my comment is about the enterprise and imaginary scaling problems, realize that those situations don't apply to some programming problems.


> Based on what's in the article, it wouldn't take much to move these files to SQLite or any other database in the future.

Why waste time screwing around with ad-hoc file reads, then?

I mean, what exactly are you buying by rolling your own?


You can avoid the overhead of working with the database. If you want to work with json data and prefer the advantages of text files, this solution will be better when you're starting out. I'm not going to argue in favor of a particular solution because that depends on what you're doing. One could turn the question around and ask what's special about SQLite.

If your language supports it, what is the overhead of working with SQLite?

What's special about SQLite is that it already solves most of the things you need for data persistence without adding the same kind of overhead or trade offs as Postgres or other persistence layers, and that it saves you from solving those problems yourself in your json text files...

Like by all means don't use SQLite in every project. I have projects where I just use files on the disk too. But it's kinda inane to pretend it's some kind of burdensome tool that adds so much overhead it's not worth it.


> what's special about SQLite

Battle-tested, extremely performant, easier to use than a homegrown alternative?

By all means, hack around and make your own pseudo-database file system. Sounds like a fun weekend project. It doesn't sound easier or better or less costly than using SQLite in a production app though.


> You can avoid the overhead of working with the database.

What overhead?

SQLite is literally more performant than fread/fwrite.


That's exactly what I was going to say. This seems more like a neat "look Ma, no database!" hobby project than an actual production recommendation.

So you trade the overhead of SQL with the overhead of JSON?

I'd say having your house fall to a third of its value while your electricity bill triples is a perfectly rational argument for opposing a data center. If the data center is that valuable, pay the residents current market value for their property and give them a million dollars to uproot their life. Letting giant corporations impose massive costs on the folks that don't have the money to buy politicians is not an efficient outcome.

I honestly don't understand Microsoft's AI strategy. It seems to be built around automating the writing process. If you ask MS 365 Copilot (as opposed to the many other Copilots) what it can do, it's deeply disappointing:

"Can you edit the Word document so the format is in line with these requirements?"

"No, but I can help you draft an implementation consistent with the requirements."

"Can you add this section to the 35 individual copies of this document in this OneDrive folder?"

"No, but I can help you draft [something]."

This is NOT the AI revolution anyone was waiting for.


This is decidedly the result of a lack of strategy. Microsoft isn’t a single unified borg.

Instead, all the little individual teams got their hands on these capabilities and they figured out where to shove it. At “best” there would have been the head of Windows or Office or whatever saying to all their reports “go do AI!”


You can run all the Office apps in a browser, and update documents that are on SharePoint live in collaboration with someone else. Maybe that's not earth-shattering, but it's quite a big change from huge separate Office legacy apps. It must have been a big effort decreed from the top. Given Microsoft leadership is obsessed with AI, you'd think they'd be pushing hard.

Atleast someone at some level need to think about how the user is going to use copilot right?

Its shocking how they didnt. Imagine how shit the culture must be when employees arent bothering to consider how the user will use the feature, just focussing on getting it through

Need? No. Should? Probably!

It looks like you're trying to write something! Click here to have me fuck it up for you!

Here is a fun one. I had a column with around 200 entries and there were some duplicates in it. I just wanted to see which were duplicates and remove some of them.

I selected the cells and asked copilot to tell me which ones were duplicated. Copilot had to ask me to copy and paste the cell contents in its chat box. It couldn't even detect which cells were selected and read them

Why even have copilot inside excel when it can't even read a cell? This is what happens when all you care is about KPI metrics or what not


When implementing an AI feature in a product recently, I noticed a tendency of management to steer towards a limited, well-behaved feature set that straight jackets the underlying model. This resulted in similar experience to what you describe. Maybe this is control and accountability thing? If I were to do this, I'd just slap a bunch of product specific tools (MCP, CLIs, HTTP API wrappers, etc) and skills (how to use those, best practices) with an agent and call it a day - if it can do more but also can fail, that's fine by me. That's why I like the idea of WebMCP more than custom built, limited AI chat interfaces that pop up everywhere nowadays. Just let Claude access everything and dump knowledge into it.

> This is NOT the AI revolution anyone was waiting for.

It's Clippy. All over again.


Don't let Louis Rossmann hear you say that, though.

Consider applying for YC's Summer 2026 batch! Applications are open till May 4

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

Search: