Hacker News: avidphantasm

New comment by avidphantasm in "AI is slowing down"

avidphantasm — Mon, 08 Jun 2026 23:19:59 +0000

The AI labs are racing to create a moat out of trillion-parameter models and the GPUs that can run them. The problem is this is the wrong architecture for most AI inference use cases. On-device inference is where this is going, clearly Apple believes this too. So Zitron is entirely correct about this AI datacenter build out being a boondoggle with no ROI.

New comment by avidphantasm in "Pentagon raised threat of Israeli spying on U.S. to highest level, sources say"

avidphantasm — Sat, 06 Jun 2026 22:34:19 +0000

They have proven to the world they have a deterrent akin to a nuclear weapon, but they can actually use it.

New comment by avidphantasm in "Coreutils for Windows"

avidphantasm — Wed, 03 Jun 2026 10:47:40 +0000

And if \ was an alias for C:\ this would just be \mountdir.

New comment by avidphantasm in "Coreutils for Windows"

avidphantasm — Wed, 03 Jun 2026 10:45:39 +0000

Or just use sane names like \\MyDivision\Share01\MyData and mount that to \Network\Share01 or some such.

New comment by avidphantasm in "Coreutils for Windows"

avidphantasm — Tue, 02 Jun 2026 22:35:24 +0000

Nonsense. You can mount filesystems to mount points in much the same way as is done in Unix. No one would ever need to do that.

New comment by avidphantasm in "Coreutils for Windows"

avidphantasm — Tue, 02 Jun 2026 22:34:00 +0000

It’s arcane and technical for no reason. /Users/ME/Documents, /Media/MyThumbDrive/…, etc. are much clearer and less confusing than C:\…

New comment by avidphantasm in "Coreutils for Windows"

avidphantasm — Tue, 02 Jun 2026 17:52:14 +0000

No, they need to ditch drive letters first. The NT kernel and NTFS don't even require them (I used to mount disks without drive letters back in the NT 4 era). They just don't care enough to get rid of this annoyance.

New comment by avidphantasm in "How Socialism Could Work"

avidphantasm — Tue, 19 May 2026 11:03:01 +0000

And how do you keep capitalists from capturing the instruments of justice and subverting them to punish their enemies?

New comment by avidphantasm in "How Socialism Could Work"

avidphantasm — Tue, 19 May 2026 10:11:40 +0000

Discusses the concept of Market Socialism, which is a hybrid system meant to avoid the worst aspects of both Socialist and Capitalist systems, while putting the goal of human fulfillment at its center.

How Socialism Could Work

avidphantasm — Tue, 19 May 2026 10:11:40 +0000

Article URL: https://confrontingcapitalism.substack.com/p/how-socialism-could-really-work

Comments URL: https://news.ycombinator.com/item?id=48191437

Points: 6

# Comments: 6

New comment by avidphantasm in "Apple Silicon costs more than OpenRouter"

avidphantasm — Mon, 18 May 2026 16:28:00 +0000

It's actually a bit faster than that now it seems, about 112 tok/sec.

Configuration:

Gemma 4 31B Instruct Q6K Context size 40960 LM Studio 0.4.13+1 Metal llama.cpp v2.14.0 LM Studio MLX (Apple M5) v1.6.0

Here are my results:

prompt eval time = 32545.36 ms / 5625 tokens ( 5.79 ms per token, 172.84 tokens per second) eval time = 20227.99 ms / 310 tokens ( 65.25 ms per token, 15.33 tokens per second) total time = 52773.35 ms / 5935 tokens

This was for interacting with a local MCP service, running a tool that returns a ~20KB text file to the agent to add to the chat context.

I'm seeing about the same number of tokens/second on an M2 Ultra that I have access to (also with 128GB of memory).

This is surely apples-to-oranges to the OP results (and I don't spend a great deal of time benchmarking these things, so my methodology might be lacking), but it's interesting seeing okay performance for a top open model. For most use, however, I find Gemma 4 26B A4B (Q6K) to be good enough (esp. for MCP calling) and much much faster (~1,200 tokens/second).

New comment by avidphantasm in "Apple Silicon costs more than OpenRouter"

avidphantasm — Sun, 17 May 2026 22:59:03 +0000

Not sure where 40 tokens per second is coming from. I’ve seen 95-100 tokens per second on M5 Max 128GB running Gemma 4 31B. I’ve done experiments where it is faster than Claude Opus 4.5 for the same prompts.

New comment by avidphantasm in "Local AI needs to be the norm"

avidphantasm — Sun, 10 May 2026 21:35:43 +0000

Ultimately, information is a public good: it is non-excludable (you can’t stop people from using it) and it is non-rival (we can all use it at the same time). Public goods are often very useful, and because they are non-excludable and non-rival, ultimately can’t have a market-based business model. I would class open-weights AI models as public goods, and would support government expenditure to produce them.

New comment by avidphantasm in "Are the costs of AI agents also rising exponentially? (2025)"

avidphantasm — Sat, 18 Apr 2026 10:25:11 +0000

Calculating hourly costs for these models makes me think that the decision of when to hire an SWE vs. increase use of AI may follow a similar pattern to the decision to use cloud compute vs. on-premises. I don’t cost $120/hr (incl. fringe), but my employer pays my salary all year long, no matter if I am working or on vacation. Whereas if they use an AI model to do the same work, they may be happy to pay $120/hr or more, since they may only use the model for a small fraction of 2080 hours per year, so they’d still save money, and not have a messy human to deal with.

New comment by avidphantasm in "We gave an AI a 3 year retail lease and asked it to make a profit"

avidphantasm — Thu, 16 Apr 2026 20:35:27 +0000

I would be very surprised if they can scale hiring contractors to reliably renovate buildings.

New comment by avidphantasm in "Darkbloom – Private inference on idle Macs"

avidphantasm — Thu, 16 Apr 2026 11:31:31 +0000

If you start buying minis, then you need to house, power, and cool them. So you are building a mini data center. If you are building a small data center, economies of scale will drive you to want to build larger and larger. However, this gets expensive and neighbors tend to not like data centers (for good reason). To me this seems like asymmetric warfare against hyper-scalers.

New comment by avidphantasm in "Backblaze has stopped backing up your data"

avidphantasm — Tue, 14 Apr 2026 10:30:55 +0000

I recently stopped using Backblaze after a decade because it was using over 20GB of RAM on my machine. I also realized that I mostly wanted it for backing up old archival data that doesn’t change ever really. So I created a B2 bucket and uploaded a .tar.xz file.

New comment by avidphantasm in "I just want simple S3"

avidphantasm — Tue, 14 Apr 2026 10:08:05 +0000

I’ve recently switched from Minio and Localstack to Garage. For my needs (local testing) Garage seems to be fine. It’s a bit more heavyweight and capable than I need now, but I like that it may give me the option of having an on-premises alternative to S3-compatible stores hosted in the cloud. The bootstrapping is a pain in the ass (having to assign nodes to storage and gateway roles, applying the new roles, etc). It would be great to be able to bootstrap at least a simple config using environment variables. However, now that I have figured out the quirks of bootstrapping, it just works (so far; again, I’m not doing anything complicated).

New comment by avidphantasm in "Parallels confirms MacBook Neo can run Windows in a virtual machine"

avidphantasm — Sat, 14 Mar 2026 11:48:50 +0000

I think the major reason for the aggressive price point of the Neo, and for not raising RAM and SSD upgrade prices in the MBP much, is that Apple is willing to give up some hardware margin to have more devices to sell services to. Unless I am mistaken, services have been key to Apple’s recent revenue growth. This isn’t a bad thing at this point, but could auger poorly if they foolishly chase recurring revenue at the expense of hardware quality (their software quality has already slipped in recent years).

New comment by avidphantasm in "Peter Thiel's Antichrist Lectures"

avidphantasm — Fri, 13 Mar 2026 12:00:41 +0000

What would China do to such billionaires run amok?