Hacker News: mv4

New comment by mv4 in "I'm in a Meeting"

mv4 — Fri, 09 Oct 2026 17:12:48 +0000

For maximum plausibility, can you please add the "scope deliverable asset" lady voice track?

New comment by mv4 in "Man discovers his parents' coffee machine used 1TB of data in 10 days"

mv4 — Fri, 09 Oct 2026 15:55:37 +0000

How likely is that the machine is compromised with malware?

New comment by mv4 in "Google Playground: Create and play custom games"

mv4 — Wed, 07 Oct 2026 16:06:06 +0000

Can't test. 403 in the US.

New comment by mv4 in "Google Playground: Create and play custom games"

mv4 — Wed, 07 Oct 2026 16:04:34 +0000

In the US, got 403.

New comment by mv4 in "Why do tech CEOs sound like doomsday cult leaders?"

mv4 — Tue, 06 Oct 2026 22:01:55 +0000

Because many of them exist on a level where they can do pretty much anything they want, and nobody ever tells them No. It messes people up.

New comment by mv4 in "Meta's Muse AI agent is building a dossier on you"

mv4 — Tue, 06 Oct 2026 21:58:02 +0000

I was not defending Meta. AI agents rely on context (your preferences, basic facts, likes/dislikes) to be useful. It is also very likely that Meta goes beyond what's really required.

New comment by mv4 in "Meta's Muse AI agent is building a dossier on you"

mv4 — Tue, 06 Oct 2026 21:35:42 +0000

Well that's something Meta is very good at. Which is why I would never use that tool.

New comment by mv4 in "Meta's Muse AI agent is building a dossier on you"

mv4 — Tue, 06 Oct 2026 21:11:58 +0000

An agent needs to keep persistent facts (memories) to be useful.

I used to work for Meta on AI infra and privacy. I built my own version that makes it clear what facts it stores about you and you can edit or remove them. It also keeps separate contexts for you and your group (or household) which the other agents don't. Testing with friends and family now. Happy to share more info if anyone is interested.

New comment by mv4 in "Meta’s Muse is an adorable privacy and security dumpster fire"

mv4 — Tue, 06 Oct 2026 19:32:32 +0000

I forgot about the downvotes! OK I will need to do a proper Show HN at some point.

New comment by mv4 in "Meta’s Muse is an adorable privacy and security dumpster fire"

mv4 — Tue, 06 Oct 2026 19:16:09 +0000

It's evident that agents still can't be trusted to transact on your behalf. Too many things can go wrong.

New comment by mv4 in "Meta’s Muse is an adorable privacy and security dumpster fire"

mv4 — Tue, 06 Oct 2026 19:10:28 +0000

Ex-Meta employee here, used to work on AI infra and privacy.

I felt every agentic product was focusing on personal accounts. So I started building a platform for small teams and households (where you can have a personal workspace and a shared workspace). I am now testing it with friends and family, so if anyone wants an invite - join the list.

https://getcinnamon.ai/

How it's different:

1) personal (private) and group context (e.g. household) 2) multiple guardrails 3) different architecture (not spinning up a VM for every user) while still protecting personal data 4) not run by Meta

New comment by mv4 in "IBM Bob"

mv4 — Fri, 04 Sep 2026 18:23:40 +0000

what happened to Watson?

New comment by mv4 in "Ask HN: Why were OpenAI, Claude, and Grok simultaneously down?"

mv4 — Thu, 03 Sep 2026 18:05:37 +0000

Gilfoyle's AI deleted all software!

New comment by mv4 in "Ask HN: Why are OpenAI, Claude, and Grok simultaneously down? Coincidence?"

mv4 — Thu, 03 Sep 2026 18:03:04 +0000

In every document created by Claude.

New comment by mv4 in "Apple caught off guard by AI demand for Mac Mini and Mac Studio"

mv4 — Mon, 31 Aug 2026 20:27:17 +0000

It's quite shocking to me how many experienced, tech-savvy people, who used to care about cookies and ad tracking - are now willingly sending their business strategies, highly confidential contracts, and intimate personal issues to a cloud provider because "it is only $0.0x per million tokens!".

New comment by mv4 in "Meta reaches $16.68B settlement over social media harms to children"

mv4 — Wed, 26 Aug 2026 16:17:10 +0000

I remember when I was working at Meta, I was offered a "high impact" opportunity to work on youth products. Nope.

New comment by mv4 in "Anthropic's best AI model struggles to attract users as cheaper tools thrive"

mv4 — Mon, 24 Aug 2026 16:46:50 +0000

Hermes is the way to go. Accounts provided by the frontier labs are just too volatile.

New comment by mv4 in "GPT 5.6 Sol is the best "vision" model OpenAI ever released"

mv4 — Mon, 17 Aug 2026 15:39:00 +0000

It 5..10 lines of code. :)

New comment by mv4 in "GPT 5.6 Sol is the best "vision" model OpenAI ever released"

mv4 — Mon, 17 Aug 2026 15:36:59 +0000

That's correct, and I was specifically referring to the example chosen - where scale and perspective are known. Template rotation is relatively easy as well - but partial obstructions would pose a problem.

Another application where template matching would work brilliantly? Car counting in parking lots using satellite imagery.

Source: I did this [1] using OpenCV and template matching. Outperformed "Cars Overhead with Context" models.

https://abcnews.com/International/satellite-data-suggests-co...

New comment by mv4 in "GPT 5.6 Sol is the best "vision" model OpenAI ever released"

mv4 — Mon, 17 Aug 2026 15:29:28 +0000

Exactly my point. Template rotation is a trivial operation as well.