Cursor seems to have intentionally dropped semantic/embedding-based codebase indexing. Their explanation is that newer coding agents are now good enough at searching repos directly. They can grep in parallel, inspect directories, read likely files, and refine the search themselves. Cursor says this now works as well or better for most cases. [1]
So the architecture shifted from something like “chunk → embed → vector DB → retrieve” to “local search index/ripgrep → agent explores.” There is still an index, but it’s a conventional local search index. No embeddings are involved in that retrieval path. [1]
The interesting part is that Cursor was still reporting a ~12.5% accuracy gain from semantic indexing in January 2026. [2] So this looks less like “semantic search never worked” and more like agentic search improving fast enough that the extra infrastructure stopped being worth it.
Grok 4.6 XHigh uses 2.52x fewer tokens per task than Fable Max. It's much more efficient. It also has low market penetration, which is why SpaceX is selling so much of their compute to Anthropic et al. From a business perspective, they're capitalising on the market very well. If Grok becomes more popular we should expect to pay more.
His word is worthless. Especially given that Gruber gave him a draft of the publication and he didn't feel the need to correct any details.
Perhaps Claude did just find the repo and reproduce it without any prompt, but also the name? Yea hmm, like 99% sure Claude would've mentioned where the majority of the project came from.
> like 99% sure Claude would've mentioned where the majority of the project came from.
I doubt that. I’ve seen it produce thousands of lines of code or text with like a 2% string distance from something available online and not attribute it. Multiple times.
You can carelessly leave employee access on the confidential and private company data and in nearly all cases depending on employment agreement it will be against the law to share that data after your employment
It doesn't sound to me like Apple is pursuing individuals that merely had company data on their computers but rather these individuals then went on to share said data with other individuals.
It works well enough if you deploy a separate version of the app for each tenant, and upgrade each tenant's app + db in lockstep.
It's a bit of a 1990s setup (as the sibling comment says, it's more or less the same model as a widely-deployed desktop app), but it does scale and can be useful in situations where strict isolation between tenants is beneficial.
Rust releases are just compiler toolchain, maybe some new syntax features. Java includes the JVM which is subject to way more security issues and needs much more frequent updating.
There are some actual removal of feature too (breaking backwards compatibility on purpose). But those come with deprecation warnings for years before the actual feature is removed. And even then quite often it is still possible to enable with some feature flag for a version or two.
Patches are released continuously. The upstream versions get them immediately and they are then backported to LTS versions. Whether the patches actually become available simultaneously I cannot say without.
The article doesn't make really controlled findings. There's an argument to be made that the increase to diseases isn't purely lack of vitamin D but lack of exercise.
Our ancestors got lots of Vit D but they also got lots of exercise while absorbing the sun.
I still don't think it's going to be wise to go out and just bake in the high UV of early afternoons but rather it's important to go outside early to mid morning or late afternoon and absorb some sun without copious amounts of sunblock.
I'm quite happy Apple stays focused on their products. They enter a market when they can own it end-to-end -- it makes no sense for them to all of a sudden become an AI chip house or AI server house.
There is a lot of money in AI chips, and Apple could definitely get a fairly large slice of that business if they put the work in (well, if they are putting the work in now, depending on what Baltra is really about).
I'd still say getting 7 years of free OS supports for hardware is pretty incredible. All-in-ones like iMacs are always going to be wasteful on the environmental side of things faster because displays will outlive the usefulness of the rest of it by a large large margin.
At least Apple does try a bit to be a responsible recycler and you can always take your old hardware to them.
I have a 2019 GPU in my windows box and I’d be pissed off if it stopped working in the next few years. Computers nowadays are useful for much longer than they used to. Spectre class vulnerabilities which take 50% performance to properly (kinda) mitigate are the only reason to maybe upgrade if all you do is browse facebook and pay bills.
7 years is generous for phones but desktop computers have had a much slower performance ramp up where even a 10 year old desktop should be perfectly cromulent for office tasks as long as you give it an SSD and enough RAM.
The sad thing is that newer macOS versions do still work really well on these unsupported machines - OpenCore Legacy Patcher proves it. All Apple would need to do is spend some time testing things, there's nothing to actually implement to make it work.
Isn't Apple supposed to be a services company now anyway, making money off of this installed base with cloud subscriptions?
How is the OS Free? It's part of what you pay for.
I have a 11 year old PC -- running Cosmos now, and it's still faster than my hobbled M4 PRO with 48 GB, work mac with all it's corp spyware cruft on it.
We should expect more than 7 years out of all our tech hardware.
EDIT: I say this as a person who spent a couple weekends trying to get various forms of Linux running on a 2017 Macbook Pro, because it was stuck on VEntura.
Free OS updates is the norm now though. If you'd bought a Windows PC 7 years ago, or used Linux, or just about any other desktop OS, you'd have had free updates.
I guess you forgot about windows TPM controversy. Comments like this make me realize a good majority have no idea what they are talking about and still decide to voice an uninformed opinion.
reply