Maybe is a weasel word when they know it isn’t. Also, they are trying to change the definition over time.

Its iterative improvement but with the same risk of hallucinations and the same difficulty in giving it access to sensitive data or sensitive authority and it messes up significantly.

Their shows are good, but it’s a really limited selection. It’s not a primary service to replace something like Netflix, with a huge (but shrinking) range of shows and movies that they didn’t make.

The best thing about open source is that a huge new entrant into the system only adds to the system. China will undoubtedly make changes but others can choose to implement or not. Some may be good for all. Some may suit Chinese users only. Some may suit government only. Some may suit Chinese government only. Whatever, it’s options and novel methods.

And the more that do this globally, the more independence and options we all have.

Enshittified is more specific.

We moved to LibreOffice

It’s an adjustment, but so is a jump between versions of office.

I think that’s part of the plan and risk of AI. They will try and rent cpu usage and knowledge. Hopefully, they don’t succeed and instead we have open source AI and less dependence on tech. Even without a crash.

As much as demand for ram and video cards is killing everyone currently, there should be an increase in supply if it’s prolonged enough. Even without a crash.

Now, add a crash and we may have better home ram and video cards for cheap. I think given American volatility, and tech enshittification, people no longer trust the cloud and are moving away from it. They are also sick of subscriptions.

People with their own community websites did this before reddit existed. Then yelp and google did it whole-scale. Then they enshittified it when those websites were gone/didn’t rank.

I don’t think the Linux foundation is the reason for Linus Torvalds saying AI is not excluded. This is not what you are claiming, but seems to be the onference of your post. The Linux foundation only spends a tiny fraction of their budget on the Linux kernel.

The whole point if open source is that anyone can contribute. Companies that see the value in Linux can donate. All these AI companies companies do so as it is in their interest to do so due the stability of Linux and its use in servers, now llms. It is the same reason nvidia are rapidly improving their Linux driver support. Web servers didn’t need great support and Linux home users were a tiny fraction of their users. So they didn’t give a toss. Then llms used graphics cards and they saw profit and opportunity. No different to these other companies, they invested where it suited them to, not where tje community wanted.

Limis is smart enough to not rock the boat with sponsorship, but not compromised enough by the reality of required sponsorship, in my view, to change ethos or dorection to chase said sponsorship. These donors, although largest, are worth much less than the thousands of free volunteers and many thousands of smaller donors that won’t contribute to a compromised project.

I think AI is a bubble and is overhyped and I expect a financial crash due to it. I also foresee it is here for good, even though it won’t be the revokution the investors hoped it to be. It is still revolutionary, but with severe flaws. I would like to be in a situation where ai continues to be best run on Linux and small home models are useful for most people rather than a dependence on cloud computing. We are seeing a transition to cloud based infrastructure for comoute, beyond what we’ve seen for storage of discrete consumer data, like photos. That is scary as once the power is gone from consumers to have alternative options from the big 7, the more the risk for society, in the same way that social media has had a terrible effect on data bubbles and personal conflict.

Phones have moved to iterative rather than revolutionary and their costs is going up rather than down. Likely people will choose to upgrade them less. Simolar for computers 10+ years ago unless you’re a bleeding edge gamer. Even then there was the introduction of cloud gaming.

What we need is to continually have consumer options that respect privacy and user rights, which Linux is best placed to do. In doing so, it should offer all available options for computation, which includes ai. As linis says, ai is a tool. It would not make sense to categorically refuse it. Code should be approved based on merit. Already, we are seeing consumer pushback on ai. Spotify, for instance, now marks content as ai. Meta too. Xitter et al. And they all tried to initially pass off slop as content. They are recognising both the poorer quality and the customer resentment about that and adjusting accordingly.

Ai money is a bubble. I’m glad that some of that rich investor money is going to a good cause before the bubble bursts. Many of those companies were already donors before the launch of chatgpt which started the rollout of ai everywhere. Hopefully it continues afterwards. I hope their donations are partly what leads to their downfall, or at least the removal of their monopoly powers.

Yes, I agree. So the choice is self host, with the risks of that but also the hassle for you and family. Or use a smaller service focused in privacy that might(will) get hacked eventually, or just hand your data over willingly.

Is that why apple sued them?

Cats don’t learn about you. They learn how to control you. Oh.

It’s only accessible by those you share a link with to your personal child(ten) gallery. They don’t share photos with you, just access your child photos to comment and like.

It also has tools for recording things like height and weight as the child grows. It’s probably possible with a locked down j stagram account but I wouldn’t trust meta to not use the data nor to not change the settings over time.

I used tiny beans with my family. It’s a micro social network designed for sharing photos with small groups, particularly of babies and kids.

I agree fully with the crap in the feed. The inmediacy though, not so much. I recall when they started that and people were still posting they tried to use it for engagement. A new post from a friend that had engagement would be repeatedly resurfaced in your feed.

It did mean less consequential posts from days earlier did feel stale and pointless when they were pushed through again, though, if that’s what you meant.

True, but a cancelled order costs you time and hassle. Leaked id and credit card info can cost much more.

Not with a shelf. It rests there. I’m not talking fluid diets, but soft not fully formed mass. There is a whole scale for assessing stool. For most people it’s unnecessary but it can be helpful for those with things like crohns disease or IBS or ibd. Or even just for descripties for people describing discrepancy to their doctor.

Yes, when ads companies did surveys to match ads with your interests, I’m like, no, irrelevant ads are even easiest to ignore.

What if you didn’t have solid turds, generally?

Oh, I assume it will also increase fraud. Lots of dodgy sites will ask for ids and/or credit card details to verify age. I wouldn’t trust them without a vpn, age verification or not. Edit: typo

Its to allow inspection of faeces prior to flush, I believe. For health purposes.