Hi all! As promised, this is a departure from the normal Sizzle format.

Please enjoy and have a ripper weekend. See you on Monday!

Mentioned in today’s edition: ChatGPT, Elon Musk, Home Assistant, Kimi K2.5, WhatsApp and Waymo robotaxis. Plus, deals on Bose headphones, Samsung tablets and Kogan prepaid plans.

The News

Let’s talk about AI. (Specifically — and from here on, referring to — generative AI).

Earlier this week, I wrote a Sizzle news item about Clawdbot, the open source, locally run AI system that people were letting run amok on their systems. The tone of the piece was pithy, noting that doing this had immediately caused some enormous security fuck-ups.

Later that day, Josh wrote on the Sizzle forum that he thought my write-up was a pretty negative take on a cool tech: “It’s not basic automation, it’s the marriage of LLMs and a local computer that hasn’t occurred yet and is a bit of a real power tool.”

I was grateful for the feedback! I thought about it, as I have been thinking about AI just about every day that I’ve written the Sizzle. I’ve been pretty upfront about my skepticism on the AI hype that’s taken over the world since ChatGPT was released in November 2022. 

There’s been immeasurable harm from its reckless rollout by powerful people primarily motivated by a desire to greedily grab as much power as they can. It instantly shredded any hope we had at a shared sense of reality, with the Trump Administration putting out AI videos. It’s a technology that is like a Dunning-Kruger supercharger, convincing people that they can do things they can’t or shouldn’t do — like the US’s acting head of cybersecurity who is accused of putting classified info into a public version of ChatGPT. Most of the products are based on mass cultural theft which, because of the carelessness, also incorporates disgusting harmful material like child sex abuse material, like Amazon just found out. It is unambiguously speeding up our planet’s climate disaster, fuelling a resurgence of fossil fuels. 

All of this is still true. But Joshua’s comment crystallised something that had been simmering away in my mind for a while: that the AI bubble, the AI industry, and the technology itself are actually separate things, even if they’re extremely intertwined.

As a tech reporter, it’s my job to actually use things to understand it. When I’ve tested or trialled AI, especially recently, I've found that a large language model’s capability to interpret and translate natural language is genuinely useful. It’s been a good supplement — not a substitute — for traditional search. It’s helped me in situations where I’ve been searching for something but I haven’t perhaps used exactly the right keywords. 

Then, a few weeks back, I wrote how I’d used it over the Christmas break while dicking around with smart home stuff that had previously flummoxed me. The Verge’s smart home guru Jennifer Pattison had a piece just this week about how she used Claude Code to vibecode a system to bring her insane set up together. 

That was a bit of an a-ha moment. Not because it worked perfectly for me now, but because I started to see one way this could go. What if I can use these tools to have more control over my own devices now? Using AI to sort out my own self-hosting, even vibecoding my own personal stuff, so I don’t need to rely on cloud providers and foreign tech companies controlled by billionaires to do stuff with my own computers in my own home?

Ah, you might say, you’re ignoring that using generative AI makes you more dependent on the companies providing the service! It’s a good point and one of my major concerns with the dynamics of a world intermediated through AI systems controlled by a few. I think it’s pretty fair to look at how someone like Elon Musk is running AI companies and his goals and think, hmm, maybe we shouldn't trust him with even more power.

Open weight LLM products are an interesting counterpoint to this. They're still pretty centralised, given the enormous resources required to put out a model, but the reality is the last few years has seen a constant stream of free LLMs that you can run locally that are only a few months behind the top tier, close weight ones. Just this week, Chinese company Moonshot put out Kimi K2.5 which is getting rave reviews. There’s something pretty cool about being able to spin up Qwen:8B on my 2023 MacBook Pro. It’s completely run on my terms, limited to what I want it to do, etc. This feels like a step towards decentralising this technology in a genuinely beneficial way.

Now, before you get concerned, I’m not writing this to announce a pivot to being an AI fanboy.

I still am fucking pissed off that arrogant tech bros have turned the entire world into science experiment about what happens when you force a new, disruptive technology onto everyone with few safeguards. 

I’m not one of those AI safety guys who is worried about SkyNet, or whatever. 

I’m actually scared and furious and devastated that we all have to deal right now with non-consensual sexual image mass production, endless slop, job layoffs, environmental destruction because of the decisions of a powerful few who have the audacity to spin it as democratising or whatever. Lord knows what’s going to happen to the economy given that, even if there are some uses for AI, I haven’t seen any indication it’s going to pay off what has been spent on building it out. 

When it comes to my work, nothing has made me think that AI can or will replace me. I personally loathe AI generated writing. It’s lazy, it sends a signal you don’t give a shit, and it’s bad. 

Thinking and feeling and communicating are the crux of being human. I don't want to give that up. It’s why you pay me $6 a month! As a reminder: the Sizzle is written by me, a human, and will always be written by a human. 

But I think that it's intellectually honest to separate the tech from those who are wielding it. It's like how I truly believe social media can be a really powerful, beneficial technology, but I don't think that big social media companies are designing it in that way. With AI, I think there’s a path to being able to ethically use this technology to make our lives better, not worse. 

The Sizzle is fundamentally a newsletter about being obsessed with technology. I love gadgets and gizmos! I don't think that — just because the industry, the culture, the world that’s been built around a technology is fucked — that I need to treat the technology itself as intrinsically cursed.

At the same time, it would be irresponsible to cover a technology completely outside of its context. I think good coverage can consider it on its own, while also making sure it's not treated as a purely hypothetical thought experiment, but a real thing that's impacting people's lives now. I want to be curious, but not credulous.

Part of understanding a technology is honestly thinking about its flaws. One of the root causes of so much of the hate towards AI is that there’s a whole industry of people who are using the tech poorly, often to the detriment of others. AI’s fundamental problem with hallucinations means it’s best used in low stakes situations, like on things that don’t matter or are being closely reviewed by a person. That’s why I feel fine about using it to set up Home Assistant, but not to run my bank account. 

To bring it back around to Josh and Clawdbot, I do regret being a bit glib and imprecise with what I wrote. Honestly? I got lazy and just looked at it and said “fuck these annoying AI-maxxing hustle bros”, even though I knew in my heart of hearts that there was a kernel of Clawdbot that was interesting there. In my defence, it’s hard not to see all the bullshit and overpromising and instinctively react against it. But I want to be who sees and writes about the world as it honestly is, as it could be, not just how I want it to be because it's an easy dunk.

So, is it cool that one guy vibecoded an (apparently poorly secured) personal assistant that people are doing some (sometimes ill-advised) interesting things with? Yes. Is it also funny that careless dumbasses are immediately fucking up their lives in their rush to automate every part of their lives with reckless abandon? I mean, obviously yes. I think both things can be true at once. 

What’s important here is understanding that there is a distinction. I want to be precise and think hard and figure out where technology is all headed. I want to be part of CHOOSING that future where we get the best out of technology, and fight the worst uses. 

So what does this mean for the Sizzle? Nothing different. I will continue to call it as I see it, to be cynical towards powerful people and bad intentions; and to spend the lead up to every Sizzle edition finding the best and most interesting things about tech to share with you.

Also, one thing I’m certain AI can’t do yet is “constantly simmer on something until it breaks and writes 1500 words on a topic, knowing full well that it’s going to make some people mad” — so I guess that’s one thing I’ve still got going for me.

Thoughts? Feedback? Pithy responses? I read all emails, comments, etc. Would love to hear your response via email or Signal. 

Leftovers

Australia:

Rest of World:

Discuss in Slack or Forum.

Oh, Also

An app to make cat cards for your Apple Wallet

Here's a fun little applet for Friday: Kitty Cards let you make cat-related "cards" for your Apple Wallet. What's the point? I don't know, it's pretty cute. Sometimes things don't have to have a point. Sorry to dog/Android people. 

Discuss in the Sizzle Slack or forum.

Bargains

Electrical & Electronics

Computing

Mobile

The End

😎 The Sizzle is written by Cam Wilson and emailed every weekday. It was created by Anthony “decryption” Agius.

🤖 We love robots at the Sizzle but this newsletter has always been and will always be written by humans for humans. Also by Aussies for Aussies — so all prices are in dollarydoos, of course.

🗣️ Have any feedback, a tip or just want to chat? Send me an email or Signal message. I promise to reply!

💬 Want to hang out with other Sizzlers? There’s a subscriber-only Slack server and forum if you want to procrastinate and chat about tech-related news.

🌐 The Sizzle is on Bluesky, Mastodon and LinkedIn if you’re feeling social.

💳 Paid subscriber looking to manage your billing info, change email address or cancel your subscription? Visit the Beehiiv customer portal.

🎁 Make someone's day and gift them a 12 month gift subscription to The Sizzle.

💔 Don’t want this any more? I won’t take it personally. There’s a unsubscribe button at the bottom of this email or here’s a guide.

🦺 The Sizzle has been tested to meet and exceed ISO 3533 standards.

Always Was, Always Will Be Aboriginal Land

The Sizzle is created on Gadigal land and acknowledges the traditional owners of country throughout Australia, recognising their continuing connection to land, water and community. I pay my respect to them and their cultures and to elders both past and present.