I don't think this goes nearly far enough, and is another example of our government latching onto a juicy trending topic and doing the bare minimum to make a positive sounding news byte out of. This should be an opt-in model for algorithmically driven content, infinite scrolling, etc. An opt out model just feels like they're compromising for the benefit of industry lobbyists on the issue.
Why do people (or at least politicians pretend to) believe that social networks are popular despite the content algos and not because of them? The reason Tik Tok blew up is because it shows you what you want to see without you twiddling with a dozen dials.
Maybe giving people what they want is not great for them. I am ok with this argument. But then "giving people control" won't change anything. Vast majority of users would rather put up with shocking (to me) levels of annoyance and expense just not to configure/fix anything themselves.
Most of my crowd seem to have left them behind because they went from a tool where I could see what my people were up to, and maybe the odd update about a band or a business I was interested in, to "Here are 50 things you have no interest in and by the way I deprioritised what your actual friends' posts".
This happened so much that most of the actual friends wandered off, making those posts even rarer. There seems to be a de-networking effect.
Should government be legislating that social networks work the way I want? Very good question, probably not. Am I still happy with this change? Broadly yes.
I’d prefer similar to Doctorow, that governments simply make scraping and republishing (anything that anyone who consents can access) explicitly legal, then alternate frontends may appear with their own algorithms.
I talked about this in my blog post [1]. I didn't expect regulators to pay attention to the issue anytime soon—let alone move against it. So kudos to the Australian government. Enough with the free-speech and free-market rhetoric: algorithmic feeds are actively poisoning our minds and should be regulated as strictly as possible.
> The issue is that much of the problematic content doesn’t violate any law but still is poison that shouldn’t be on anyone’s feed. The easiest example is painting feminism as some evil propaganda against men, and how women’s place in the kitchen is natural order of things. Such views have gained steam in the recent times purely because of social media (algorithms).
> The ideal regulation would be to ban or heavily regulate algorithmic feeds. If you don’t follow someone, or a page, or an interest, there’s no reason for you to get that stuff on your timeline. Only things that should be shown is what you explicitly choose. And if it does, it should be easy enough to nuke it from your timeline.
This reminds me of the YouTube creator Technology Connections trying to engage with people on bluesky, which already provides feed customization and even lets you write code to build your own feed algorithms. He loosely coined the term "algorithmic complacency" to explain how users didn't like bluesky specifically because they had to explicitly choose who or what they wanted to follow, his fear being that especially young people would rather be served slop than decide for themselves what content they want to see.
The war on drugs was a failure, and the most successful cases (other than harsh punishments while avoiding massive drug use in the first place, like Singapore) involve legalization.
The USA has a choice between two parties - democrat or republican. China only has one party, Russian has (in principle) one party and North Korea only has one dictator.
So if these places got a second party or dictator, it would be a 50% improvement.
The feed options sound good in principle, although I expect the implementation to be deliberately hobbled. I've tried Facebook's existing (well hidden) chronological feed and it turns out there's so much spam in local groups that I found it unusable.
The bigger issue is the toxic "papers please" identity verification will result in a lot fewer "Australians" on the internet, who will now show up from various VPN endpoint countries, so presumably won't get the algorithm option.
Interesting scenarios that could happen in next few years
* American internet companies start to make the judgement that supporting EU/ANZ countries with difficult legal/tax requirements is a net cost and stop doing it, just IP-blocking those countries for the lowest legal risk
* Alternatively, American govt takes the position of "American companies don't have to follow the laws of random countries just because they accept IP packets from those countries". American companies eliminate their official business presence in onerous legal regimes but keep serving users there
* If above happens, this forces the resolution of the question "how does internet jurisdiction work across borders if USA doesn't play along", and a plausible answer is that internet messages start to be treated like physical imports and nation-level internet censorship moves under the purview of customs agencies
The globally routable internet is a very special and valuable thing, but it's a historical aberration and I suspect not very stable long-term. Even today, with the internet way more locked down than 20 years ago, it's still extremely libertarian compared to almost any other economically significant venue. In equilibrium, the state will demand its cut
> American internet companies start to make the judgement that supporting EU/ANZ countries with difficult legal/tax requirements is a net cost and stop doing it, just IP-blocking those countries for the lowest legal risk
Productivity spikes, youth mental health issues trend downwards and the nutcase right-wing parties struggle to maintain momentum.
I think this is a nice idea, but three koala dollars says that in a year we find out most Australians switched back to the dopamine feed after a day or two. Nice to have the choice, though.
Maybe giving people what they want is not great for them. I am ok with this argument. But then "giving people control" won't change anything. Vast majority of users would rather put up with shocking (to me) levels of annoyance and expense just not to configure/fix anything themselves.
Most of my crowd seem to have left them behind because they went from a tool where I could see what my people were up to, and maybe the odd update about a band or a business I was interested in, to "Here are 50 things you have no interest in and by the way I deprioritised what your actual friends' posts".
This happened so much that most of the actual friends wandered off, making those posts even rarer. There seems to be a de-networking effect.
Should government be legislating that social networks work the way I want? Very good question, probably not. Am I still happy with this change? Broadly yes.
"Albanese" is the second name of the Australian Prime Minister
In Albania, Albania is called Shqipëria.
> The issue is that much of the problematic content doesn’t violate any law but still is poison that shouldn’t be on anyone’s feed. The easiest example is painting feminism as some evil propaganda against men, and how women’s place in the kitchen is natural order of things. Such views have gained steam in the recent times purely because of social media (algorithms).
> The ideal regulation would be to ban or heavily regulate algorithmic feeds. If you don’t follow someone, or a page, or an interest, there’s no reason for you to get that stuff on your timeline. Only things that should be shown is what you explicitly choose. And if it does, it should be easy enough to nuke it from your timeline.
[1]: https://shubhamjain.co/2026/02/25/algorithmic-feeds-need-to-...
https://youtu.be/QEJpZjg8GuA
“Drug dealers are now required to give their customers the option to purchase salad instead of cocaine”.
The USA has a choice between two parties - democrat or republican. China only has one party, Russian has (in principle) one party and North Korea only has one dictator.
So if these places got a second party or dictator, it would be a 50% improvement.
The bigger issue is the toxic "papers please" identity verification will result in a lot fewer "Australians" on the internet, who will now show up from various VPN endpoint countries, so presumably won't get the algorithm option.
* American internet companies start to make the judgement that supporting EU/ANZ countries with difficult legal/tax requirements is a net cost and stop doing it, just IP-blocking those countries for the lowest legal risk
* Alternatively, American govt takes the position of "American companies don't have to follow the laws of random countries just because they accept IP packets from those countries". American companies eliminate their official business presence in onerous legal regimes but keep serving users there
* If above happens, this forces the resolution of the question "how does internet jurisdiction work across borders if USA doesn't play along", and a plausible answer is that internet messages start to be treated like physical imports and nation-level internet censorship moves under the purview of customs agencies
The globally routable internet is a very special and valuable thing, but it's a historical aberration and I suspect not very stable long-term. Even today, with the internet way more locked down than 20 years ago, it's still extremely libertarian compared to almost any other economically significant venue. In equilibrium, the state will demand its cut
Productivity spikes, youth mental health issues trend downwards and the nutcase right-wing parties struggle to maintain momentum.
Can we have this please? Pretty please?
Theres alot of stuff turning up in my feeds lately that i'd rather not see.
I wouldnt be surprised if most people opt back in. But its better than an outright ban.