Oh I'd say its real. Years ago I worked as that senior who divvied up the work to a team of 4 mid-level engineers in that exact fashion (they were mostly Latin American contractors with limited english). We actually shipped some good stuff too!

Unfortunately, looking back, I could easily see Claude replacing 3 out of 4 of those developers. Myself + one other dev + AI would probably would've shipped a bit more a little faster. With that said tokens aren't free so the net cost savings would've been 2 dev salaries maximum.

I'm there now. I have a team of low level engineers in Hyderabad and Bengaluru which can be useful but I have to spend a lot of time writing very detailed specifications and instructions to walk them through what has to be done. There are plenty of instances I can point to where it would have been faster for me to implement the solution myself, by hand, no Claude. Now bring Claude into the mix and I do not really have any use for these engineers at all aside from providing operational support during IST.

We're not too much an AI-friendly company yet, so we don't officially have Claude access, but I can tell with certainty that several of my Indian teammates are using Claude anyway. I'm pretty familiar with the style of code it creates. And the comments are decidedly better English, which is a big giveaway itself.

I feel like PA-RISC actually landed with a handful of useful somewhat-complex instructions. It always struck me as the best architecture from that era, making the correct trade-off of avoided microcode but adopted stuff like completers and shift-and-add operations to minimize instruction count and maximize work done per pipeline.

Yea, there were things like shift and add to do multiplication more efficiently, but they were all one-cycle instructions. It was very regular that way, all designed around a clean pipeline without a stalls or bypassing. But, consequently, it originally didn’t even have integer multiplication or division, just shift and add and a “divide step” that you could repeat/loop. Looking at the instruction set just now, I chuckled at how simple it was. It makes RISC-V look complex.

See: ftp.parisc-linux.org/docs/arch/pa11_acd.pdf

Edit: actually, it did have fixed point multiply via the floating point unit (opcode XMPYU). No fixed point divide, though.

One level above junk. Yikes.

The problem with Oracle is how much utterly and directly depends on them. They still are the authoritative source on "what is Java" and stuff like SAP directly depends on Java and in many cases Oracle Database (the supported alternatives are AFAIK IBM's mainframe stuff and MS SQL, but I only ever saw Oracle whenever I had the misfortune of having to dig into SAP).

If Oracle goes down, a lot of businesses are going to be in a lot of trouble, and there will be a lot of fighting over the scraps.

SAP has largely switched over to their own HANA database for new deployments. Even before that their "AnyDB" database worked with pretty much any backend and is still supported. The reason you see oracle has less to do with what SAP wants and more to do with "It's what the company/engineers have already used/set up in the past".

Regarding Java, SAP has long switched over to openjdk (though oracle java is supported in some cases). Specifically, they have their own custom fork of openjdk "SapMachine" precisely to be independent from oracle (FYI: if you look at openjdk upstream you will see many core contributors are employed by SAP)

In general, I would say that the actual tech companies that have to interact with oracle's stuff are generally very well equipped to drop them or have already done so. The bigger issue are legacy "consumers" of oracle (e.g. if Pepsi has their data in oracle DB), which often do not have such contingencies.

It's rare for a large company to simply close down. Usually someone buys it out of bankruptcy, and keeps the profitable parts going. Database licensing definitely makes money.

yeah, like my university's whole accounting system...

oh no, not the hierarchy of consulting! the horrors. Better grab your nearest gpu and open model to replace them.

Disagree: The era of "<Tech Giant> will eat your startup by throwing capital/devs at it" are over. Anthropic, OpenAI et al are under too much competitive pressure to chase every quixotic product idea.

This is more about the CHIPS act than the tariffs.

If Trump sat on his hands for four years, he'd have been the best President ever.

But that wouldn't have made him, his family, and his cronies billions of dollars.

More usually expressed as "if he only played golf".

There's actually a housing shortage in the US so it's not like this is an architecture that is responsive to consumer or cultural preference and demand - local zoning laws do not even permit such freedom.

Lots of people blame Zuckerberg, but my own view aligns with the author in that much of this is falls on Alexandr Wang's shoulders (Scale AI's founder). It's perhaps somewhat ironic that the "MEI" guy (Merit, Excellence, Intelligence) was permitted to poach high-performing subject matter experts from key engineering orgs and reassigned them to data labeling - something that, let's be honest, is not where you want to allocate your top performers at an org like Meta.

This is one of those things where a (tech) celebrity founder was permitted to blew up a high-performing engineering culture. If shareholders knew the nuances of this they'd demand his ouster. His leadership has been lacking in merit, excellence, and intelligence.

Zuck hired him and has always been super hands on. He is responsible.

To an extent that's true, but while Zuck might be hands on but he doesn't typically micromanage his lieutenants. Alexandr was clearly in over his head, has never run anything resembling a high quality engineering org, and blew it up in less than a year. I think his days are numbered.

Yes, but hiring someone with no experience running an AI lab to run your AI lab is a terrible decision for CEO to make in the first place.

He does not micromanage, but at the same time he does not tune out. He has to see what’s going on.

Newspapers have said Zuck has a desk in the meta super intelligence lab

> was permitted to poach high-performing subject matter experts from key engineering orgs and reassigned them to data labeling

I think the main proponent was Bosworth, not Wang

I would say the claim that AI is going to replace most white collar work a very snake-oily term. The technology behind it however is very compelling and interesting.

I don't know enough about most white collar work to make any predictions. But I know a lot about software development and information technology because I've been a professional since 1995. The claims being made about AI's impact on that profession do not seem at all snake-oily to me.

I'm not sure if this was Flux, but one of those AI EDA tool companies had a somewhat absurd ad where the narrator stated the AI tool told them a capacitor was being used to block DC, and that's something they never learned while getting an EE degree. Now, I don't have an EE degree, but I feel like how capacitors interact with AC and DC are sort of "passive components 101" that even hobbyists learn quite early on.

The EDA space doesn't strike me as being anywhere near as SWE when it comes to AI.

Might be aimed at people who spec out the $5100 Macbook Pros with M5 Maxes and 128GB.

definitely! it has the advantage that it can run CUDA kernels but on the other hand it has lower memory bandwidth and probably loses a token/s fight for many LLMs.