Hacker News: fh9302

New comment by fh9302 in "The Israeli spyware firm that accidentally just exposed itself"

fh9302 — Mon, 16 Feb 2026 13:51:56 +0000

I would be careful with such allegations. The emails show Ehud Barak first visited the island in 2014, while Virginia claimed the rape happened on the island in 2002.

New comment by fh9302 in "Even Apple cannot explain why we need AI in our lives"

fh9302 — Wed, 19 Jun 2024 17:20:19 +0000

Apple Intelligence is run on-device and on Apple's Private Cloud Compute, it does not use OpenAI. Apple also allows ChatGPT (and later other models) to be called directly from Siri and writing tools, in cases where Apple Intellgence can't solve the task. That is for example for when some text should be composed, like a story or recipe. Nothing about this contradicts with what I have written previously.

New comment by fh9302 in "Even Apple cannot explain why we need AI in our lives"

fh9302 — Wed, 19 Jun 2024 15:37:17 +0000

Apple Intelligence doesn't use OpenAI at all. Siri and writing tools can tap into ChatGPT to compose text, e.g. create a story but that requires approval from the user. Claiming that all Apple did was "hastily shoved together selection of OpenAI API calls" is a misrepresentation of what Apple showed.

New comment by fh9302 in "Apple Intelligence for iPhone, iPad, and Mac"

fh9302 — Mon, 10 Jun 2024 20:11:46 +0000

Apple does not take screenshots every couple seconds, unlike Microsoft. That's what people were bothered about.

New comment by fh9302 in "Microsoft will switch off Recall by default after security backlash"

fh9302 — Fri, 07 Jun 2024 19:31:05 +0000

I did read the article. The person I'm replying to claims the entire debate was "uninformed hysteria", which means they thought the previous security model already required admin.

New comment by fh9302 in "Microsoft will switch off Recall by default after security backlash"

fh9302 — Fri, 07 Jun 2024 18:21:25 +0000

It is possible to access to Recall database without admin access.

https://x.com/GossiTheDog/status/1798832390070276500

New comment by fh9302 in "AMD Unveils Ryzen 9000 CPUs for Desktop, Zen 5"

fh9302 — Mon, 03 Jun 2024 10:56:33 +0000

+35% is for a single cherry-picked Geekbench AES subtest. AMD did not show the overall Geekbench improvement.

New comment by fh9302 in "Recall: Stealing everything you've ever typed or viewed on your own Windows PC"

fh9302 — Sat, 01 Jun 2024 00:05:22 +0000

Recall can't be disabled during the Windows setup. It has to be disabled manually in Windows settings.

https://x.com/tomwarren/status/1796681578984182066

New comment by fh9302 in "What's New in Neovim 0.10"

fh9302 — Thu, 16 May 2024 15:02:33 +0000

For some reason vim-airline in combination with neovim 0.10 has really bad performance when scrolling large files. This doesn't happen with neovim 0.9 and regular vim.

New comment by fh9302 in "Android's theft protection features"

fh9302 — Thu, 16 May 2024 01:10:14 +0000

It can't. This user seems to be confused about what lockdown mode is.

New comment by fh9302 in "Android's theft protection features"

fh9302 — Thu, 16 May 2024 01:09:16 +0000

Lockdown mode is used to protect journalists or other people against malware like Pegasus. It doesn't get activated by being in an unusual location, it has to be manually activated in settings.

New comment by fh9302 in "M4 near 4000/15000 in Geekbench 6"

fh9302 — Mon, 13 May 2024 04:13:34 +0000

I would recommend reading the Geekbench 6 internals document, they explain the rational behind the change.

> Geekbench 6 uses a “shared task” model for multi-threading, rather than the “separate task” model used in earlier versions of Geekbench. The “shared task” approach better models how most applications use multiple cores.

> The "separate task" approach used in Geekbench 5 parallelizes workloads by treating each thread as separate. Each thread processes a separate independent task. This approach scales well as there is very little thread-to-thread communication, and the available work scales with the number of threads. For example, a four-core system will have four copies, while a 64-core system will have 64 copies.

> The "shared task" approach parallelizes workloads by having each thread processes part of a larger shared task. Given the increased inter-thread communication required to coordinate the work between threads, this approach may not scale as well as the "separate task" approach.

Nothing about this is biased towards Apple. GB6 simply scales worse with more cores due to increased inter-core communication requirements.

https://www.geekbench.com/doc/geekbench6-benchmark-internals...

New comment by fh9302 in "M4 near 4000/15000 in Geekbench 6"

fh9302 — Mon, 13 May 2024 03:21:40 +0000

Can you share evidence of "heavy bias" towards Apple in GB6?

New comment by fh9302 in "M4 near 4000/15000 in Geekbench 6"

fh9302 — Mon, 13 May 2024 02:43:47 +0000

If your compare the Geekbench 5 results there is an approximately 8% IPC improvement. Geekbench 5 does not use SME.

New comment by fh9302 in "Apple introduces M4 chip"

fh9302 — Tue, 07 May 2024 16:52:23 +0000

The M3 Pro had some downgrades compared to the M2 Pro, less performance cores and lower memory bandwidth. This did not apply to the M3 and M3 Max.

New comment by fh9302 in "Keeping your data from Apple is harder than expected"

fh9302 — Thu, 04 Apr 2024 10:42:02 +0000

For example when you set a timer each day at 8 pm, this data is used to suggest you a timer shortly before 8 pm. It's a convenience feature.

New comment by fh9302 in "Keeping your data from Apple is harder than expected"

fh9302 — Thu, 04 Apr 2024 10:33:53 +0000

This article is highly misleading, making it sound like Siri is collecting data from apps and sending it to Apple. This is not the case, Siri Suggestions are fully on-device, though they can sync accross devices with mandatory E2EE. Apple never gets access to any of this data.

New comment by fh9302 in "Commission opens non-compliance investigations against Alphabet, Apple and Meta"

fh9302 — Mon, 25 Mar 2024 12:01:33 +0000

Safari is not hardcoded to the first position, it is fully randomized.

New comment by fh9302 in "Microsoft Launches Surface Laptop 6 for Business and Surface Pro 10"

fh9302 — Fri, 22 Mar 2024 03:53:05 +0000

Passmark is known to be a bad benchmark, something like Geekbench which has results that closely match with the industry standard SPEC would be a better comparison for real life performance.

M3: https://browser.geekbench.com/v6/cpu/5411370

Intel 165H: https://browser.geekbench.com/v6/cpu/5387822

Around the same multi-core results, significantly better single core for the M3.

New comment by fh9302 in "Bypassing Safari 17's advanced audio fingerprinting protection"

fh9302 — Sun, 10 Mar 2024 13:30:15 +0000

Do you have something more recent than a leak from over 2 years ago that has long been fixed? I'm curious why iCloud Private Relay is theatre at the moment.