Hacker News: ABS

New comment by ABS in "Apple Foundation Models"

ABS — Mon, 15 Jun 2026 13:29:08 +0000

for the on-device model, yes it runs on the Neural Engine (at the moment) so a newer chip means faster, cheaper local inference. For the server side path this Claude package is about your machine is irrelevant since it's a network call. The same API covers both, so "best machine for AI" only bites when the session is actually local.

But we can imagine that the balance of what's on-device vs what's remote will move continuously towards the former as time, improved HW and improved local models keep progressing

New comment by ABS in "Apple Foundation Models"

ABS — Mon, 15 Jun 2026 13:26:12 +0000

it's cloud, the doc is explicit that requests go straight to api.anthropic.com with Apple not in the way.

so Claude via FM dies offline while Apple's on-device SystemLanguageModel (the ~3B one) keeps working. It isn't a hybrid really: the framework just has both implement the same LanguageModelSession protocol so "local 3B" and "remote frontier model" become a one-argument swap.

IMHO what's worth internalising is that the two share an API but nothing else: the on-device path runs on Apple's Neural Engine and costs battery (you can watch ANE power ramp while it works) while the cloud path costs API credits/tokens and does zero local compute. Same code, opposite cost model.

On-device AI is a margin decision

ABS — Wed, 10 Jun 2026 18:04:53 +0000

Article URL: https://ziraph.com/blog/on-device-ai-margin-decision

Comments URL: https://news.ycombinator.com/item?id=48480247

Points: 3

# Comments: 0

New comment by ABS in "Claude Fable 5"

ABS — Tue, 09 Jun 2026 17:18:21 +0000

also: Fable takes 2× the usage of Opus

New comment by ABS in "Apple Core AI Framework"

ABS — Tue, 09 Jun 2026 11:23:10 +0000

something I haven't seen highlighted anywhere yet, while I find it very interesting, is the distributed inference across Macs (JACCL over Thunderbolt 5), an OpenAI-compatible mlx_lm.server, agentic-on-Mac.

Apple keeps MLX (bring-your-own-weights) separate from Foundation Models / Core AI.

New comment by ABS in "Apple Core AI Framework"

ABS — Tue, 09 Jun 2026 11:05:41 +0000

looks to me like the docs don't give a feature-parity table, but they do draw the "role" lines once you read across them:

- Core ML narrows to classic, non-neural ML (its own docs now point you there for "decision trees or tabular feature engineering")

- Core AI takes neural nets and transformers (the new .aimodel format, the new profiler)

- MLX stays the separate bring-your-own-weights track (its WWDC sessions draw no line back to Core AI at all)

coreai-opt is the successor to coremltools on the optimization side.

New comment by ABS in "Apple rebuilt its on-device AI stack at WWDC 2026"

ABS — Tue, 09 Jun 2026 08:51:19 +0000

among the many things I found that were not highlighted anywhere (that I could find) is the distributed inference across Macs (JACCL over Thunderbolt 5), an OpenAI-compatible mlx_lm.server, agentic-on-Mac.

Apple keeps MLX (bring-your-own-weights) separate from Foundation Models / Core AI.

Apple rebuilt its on-device AI stack at WWDC 2026

ABS — Tue, 09 Jun 2026 08:16:05 +0000

Article URL: https://ziraph.com/blog/apple-on-device-ai-wwdc-2026

Comments URL: https://news.ycombinator.com/item?id=48458183

Points: 2

# Comments: 1

Why I care so much about energy per token

ABS — Sun, 07 Jun 2026 19:50:14 +0000

Article URL: https://ziraph.com/blog/energy-per-token-vanity-sanity-reality

Comments URL: https://news.ycombinator.com/item?id=48437925

Points: 1

# Comments: 0

New comment by ABS in "Apples to Apples: MLX vs. Llama.cpp for Gemma 4 12B on an M1 16GB"

ABS — Fri, 05 Jun 2026 16:43:40 +0000

lol, it took me 48 hours to do (and re-do, and re-do) this test + write it up and now that I convinced myself to stop changing bits and just publish it... Google's just announced the Gemma 4 QAT models :-D

It would not change the core of my article since the bottleneck remains the memory bandwidth on the old M1 16GB though

Apples to Apples: MLX vs. Llama.cpp for Gemma 4 12B on an M1 16GB

ABS — Fri, 05 Jun 2026 16:36:21 +0000

Article URL: https://ziraph.com/blog/apples-to-apples-mlx-vs-llama-cpp-gemma-4

Comments URL: https://news.ycombinator.com/item?id=48414924

Points: 6

# Comments: 1

Roger Penrose is 94 [video]

ABS — Fri, 08 Aug 2025 13:41:19 +0000

Article URL: https://www.youtube.com/shorts/9TXgc_BCsjw

Comments URL: https://news.ycombinator.com/item?id=44836822

Points: 1

# Comments: 0

Three weeks after acquiring Windsurf, Cognition offers staff the exit door

ABS — Wed, 06 Aug 2025 08:00:28 +0000

Article URL: https://techcrunch.com/2025/08/05/three-weeks-after-acquiring-windsurf-cognition-offers-staff-the-exit-door/

Comments URL: https://news.ycombinator.com/item?id=44809084

Points: 9

# Comments: 3

New comment by ABS in "Why English doesn't use accents"

ABS — Mon, 07 Jul 2025 13:51:02 +0000

not obvious at all when every sentence uses "you" to indicate a general rule that applies to every Italian rather than "I" to indicate a personal experience

New comment by ABS in "Why English doesn't use accents"

ABS — Mon, 07 Jul 2025 13:03:09 +0000

you are making the mistake of confusing your experience, which is of course legit but anecdata, with "how it works" in general.

I'm an almost 50 years old Italian so not a spring chicken but I definitely learnt Italian growing up, not a dialect, and not "from school".

I guess it's the difference between growing up in a city vs a village.

New comment by ABS in "Microsoft to Cut 9k Workers in Second Wave of Major Layoffs"

ABS — Thu, 03 Jul 2025 06:59:13 +0000

there is no lost love between me and Wall Street "bros" BUT it's not their job to run the companies, their job is to buy and sell shares based on results and potential future results.

When a company decides to go public and take their money it's still the company's responsibility to run their business and keep the people they took money from in exchange for shares happy.

New comment by ABS in "Group of investors represented by YouTuber Perifractic buys Commodore"

ABS — Sun, 29 Jun 2025 06:54:51 +0000

as usual it's a matter of bubbles: most of those you know might hold such resentment, most of those I know are in fact in love with the possibilities and are trying their hardest to leverage the new tools.

My 64yo "non digital" graphic artist aunt holds a very high level of resentment to the digital ones, while most of her old friends and ex-colleagues who embraced digital way back when are stil active in the space and happy one way or another, she is not.

But she is happy she can now get near-real-time two-way translation to/from languages she doesn't speak and is also happy to bury her head in the sand when I point out that's thanks to the same tech that will have an impact on the people who do simultaneous translations as a job.

Oakley Meta Glasses, a New Category of Performance AI Glasses

ABS — Sun, 22 Jun 2025 06:32:55 +0000

Article URL: https://about.fb.com/news/2025/06/introducing-oakley-meta-glasses-a-new-category-of-performance-ai-glasses/

Comments URL: https://news.ycombinator.com/item?id=44344401

Points: 1

# Comments: 1

Microsoft advances quantum error correction with a family of novel 4D codes

ABS — Fri, 20 Jun 2025 17:26:59 +0000

Article URL: https://azure.microsoft.com/en-us/blog/quantum/2025/06/19/microsoft-advances-quantum-error-correction-with-a-family-of-novel-four-dimensional-codes/

Comments URL: https://news.ycombinator.com/item?id=44329914

Points: 6

# Comments: 0

New comment by ABS in "AOSP project is coming to an end"

ABS — Thu, 12 Jun 2025 09:49:17 +0000

FWIW: Google says Android Open Source Project not being ‘discontinued’ amidst Pixel change impacting custom ROMs

https://9to5google.com/2025/06/12/android-open-source-projec...