_{^{(From Toronto to Substack)}}

About a month ago IEEE Spectrum magazine published an online piece by Matthew Smith entitled “Your Laptop Isn’t Ready for LLMs. That’s about to change“

In the article Matthew laments that, “for the average laptop that’s over a year old, the number of useful AI models you can run locally on your PC is close to zero. This laptop might have a four- to eight-core processor (CPU), no dedicated graphics chip (GPU) or neural-processing unit (NPU), and 16 gigabytes of RAM, leaving it underpowered for LLMs.“

🤔 “That’s odd,” I thought to myself. “It sure seems like I’ve been using considerably more than ‘close to zero’ useful models on my setup.”

For comparison, I’m running a dual-core (multi-threaded) system with 128MB integrated Intel UHD graphics, definitely no NPU, and by modern standards a measly 8 gigs of RAM. The machine is about 3 years old and it was already a “budget-friendly” laptop back when I got it. As a gaming machine in 2004 it would’ve been pretty badass. Today, not so much.

Admittedly, most of the models I run locally are not (by modern standards) considered large but they’re pretty much on par for my daily needs. There appear to be a good variety of minimal desktop models to choose from and although they’re not all used for interactive chat, within my personally limited specs the number of choices is still quite large.

While Matthew makes mention of the Small Language Models that I employ, his only criticism is that these models “either scale back these features or omit them entirely“ without actually defining what “these features“ are (unless the ginormous size of LLMs is considered a “feature“?)

I’ll grant that generating responses on my hardware is noticeably slower than when using larger (remote) models but that just means that my (fully local) agentic sidekick needs to wake up a bit earlier in the morning in order to complete its high-priority tasks before my first coffee of the day. After that there are plenty of assignments that it can accomplish in the background while I finish another high-quality, fullscreen mission in “Psi-Ops: The Mindgate Conspiracy”.

All told, a 3-to-6 billion parameter model is probably the upper limit for my setup but even then I’ve got some great options like Google’s Gemma, Microsoft’s Phi, or Alibaba’s Qwen. All three come in a variety of quantized flavours that include thinking/reasoning and integrated software tool use.

If I want to use a model that’s not specifically trained for out-of-the-box tool use I can provide it with programmatic rules, not unlike how llama.cpp operates. Moreover, I can comfortably use these models concurrently with other, smaller, and more specialized models for tasks like computer vision, speech, etc.

Should I need to tighten my resource belt I can hot-swap down to slimmer language models like Liquid AI’s LFM or IBM’s Granite. Additionally, there are many derived and tweaked models available for deeply “underpowered” machines like mine.

Point being, I think that Mr. Smith got it wrong on this one. Laptops like mine are more than sufficient to run modern (albeit smaller), models. Even geriatric machines and browsers can contribute to the effort — depends on your requirements and your ability to split up the workload.

For example, there are certain tasks like generative image and video creation that my setup can’t reasonably handle but for these cases either me or my agentic buddy can farm the work out to a public interface like Google’s Colab.

There are limits, of course, but fully local agentic natural-language AI, as of late 2025, can definitely help with some of the day’s heavy lifting. In conclusion, Mr. Smith, I must judge your information to be a smidge out of date.

P.S. Regular TCL readers may recall a live example of how even browsers can run (very) limited models.

^(larger)

As of this post, Google Street View’s only capture of Housey Street at approximately this spot is from 2018. One step in the opposite direction and you’re back in 2009. Besides the interesting visual contrasts I’m curious about why the Street View car appears to be avoiding this street. Consider that there are more recent captures at each of the roads that connect to Housey yet the small avenue itself hasn’t been fully traversed by Google since 2009 (and even that “traversal” is arguable).

In Google’s latest transparency report there was a minor note about how Passport Canada had requested to ban a YouTube video featuring a passport being pissed on and flushed down the toilet. I’ve tried looking for the video but can’t seem to find it, suggesting that maybe the government had used other means to have the offending material removed.

I can only imagine that the government would’ve use the “passports are government property” excuse as a basis for this, yet considering the fact that a passport is required to both leave and re-enter Canada, to claim that they can control a person’s actions using a passport as a threat (if they can have a video banned, why stop there?), this would go directly against the Canadian Charter of Rights and Freedoms Mobility Rights clause as well as the Fundamental Freedoms portion. And since this is the highest law in the land, this is a pretty clear-cut example of the government breaking the fundamental laws governing the country, or at least coming very very close. Will anyone be held to account for this? How about a mild reprimand? Maybe a stern nod?

Originally posted at: http://patrickbay.ca/blog/?p=4081

Well now I’ve gone and done it. I’ve brazenly ripped off yet another set of photos from the Toronto Archives.

That’s what happens when I’m knackered and I’ve spent most of the night swooping around the financial district in Street View instead of posting! Oh well, if you’ll forgive a few sentences that don’t quite seem to make sense, we’ll be able get through this.

Man, I wish Google never came out with Street View. Between every few words … I manage to travel another couple of blocks.

It’s been thirty minutes since my last sentence!

Back in the olden days they wouldn’t have had these frivolous time wasters. The men with their genitals exposed to the raw atmosphere had enough to occupy their time.

The Royal Alex isn’t one of those locations that’s changed much. Guys are still freezing their nuts off out there.

The next random location — just because it happens to be a few steps down the street doesn’t mean it’s not random! — the next random location I decided on for comparison was the Elephant and Castle across the street from Roy Thomson Hall. In this case, there was a world of difference between then and now.

I figured I’d include something in the picture that I could match to something in the archives. Luckily, I was right. It’s the black building in the above photo that stands in the spot where the building below stood.

Hahaha! What the hell was that?! No one’s getting drunk in that podunk town!

… Continue Reading

The streets are a wonderful place for spotting memorable people.

The guy across from me on the streetcar this morning had the face, bristle, and thick-rimmed glasses of Elvis Costello, the fashion sense of Paul Giamatti, and the hair of Sideshow Bob. I walked through the entrance to my building with the spitting image of a female manservant Hecubus. We both passed the building’s property manager who bears more than passing resemblance to Dr. Evil.

Shouldn’t people like this be on camera? Some people don’t think so.

In an article today, The Star published a publicity piece for a group (led by Ryan Ringer) calling themselves Methinks Presents, which if you ask me, is a total misnomer.

What they intend to do is to swarm the Google Street View car that has recently been making the rounds in Toronto in order to bring attention to the “creepy nature” of Google’s project. As part of the event, they’ll probably be taking pictures in a public location.  In the process, they’ll probably be capturing the numerous random faces of people who just happen to be passing by and won’t have any say in  (or even knowledge of), being photographed. Not to mention the number of random webcams, camera phones,  and “security” and traffic cameras that cling to every available nook and cranny downtown. It’s a safe bet that everything will find its way onto Flickr, YouTube, blogs, etc.

To argue that Google is invading our privacy from the inside of a car, from a public road, means that it shouldn’t be legal for anyone to take photographs from anywhere, of anything, for any reason. Or is it just Google because they’re “evil”? Maybe Methink’s protest is intended to be somehow artistically ironic? Somehow, methinks not.

My shitter being equated to the middle of my street throws the notion of “reasonable expectation of privacy” out the window. Everything would be considered private (if the street would, what wouldn’t be?). Recorded images of any kind would have to be illegal, probably forcing the government to ban the use of cameras. While at it, why not extend the same courtesy to audio recordings? That would really suck for quite a few people.

I suppose one alternative would be to ask permission whenever you took a picture; permission of anyone in the shot (or blur them out); permission from the owners of any properties in the frame (or blur them out); permission from owners whose pets appear in photographs (or blur them out). God help you if an identifiable airplane or bus happens to pass into your shot.

Sounds silly, doesn’t it?

Not only does Methink’s plan sound horribly illegal (“hey, let’s go swarm a car because we don’t like what it’s doing”), but they’re pushing an idea that is contrary to the public good. Mine especially. I bought a brand-spanking new camera not too long ago and I don’t want to be  ambushed by Methink’s grouptards for taking a picture of the Eaton Centre.

I don’t think most people would be bothered  if they saw themselves walking down the street in a Street View scene, unless maybe they were caught doing something questionable. In that case, may I suggest maybe not doing that in public?

Oh, and when the Google car does approach, I think there’s a much better way to deal with one’s public image. Do a quick straighten-up, put on a giant shit-eatin’ grin, and give a crazy big thumbs-up as the car passes. The virtual tourist will find Toronto to have very inetersting people. Isn’t that much more productive?