We have seen over and over again chemicals which are "safe to consume" or "not that bad" actually do have very severe effects. It's just very hard to link cause and effect. When someone dies of cancer we can't pinpoint it to coming from the pesticide on the blueberries they ate a few months ago.
We have all these terrible illnesses that we ascribe to bad luck, and then all of these new chemicals we haven't fully studied yet being sprayed on everything.
Even things we know for certain aren't "safe to consume" are harmless at small enough doses. If I drink chlorine at 1000 PPM it's going to kill me, but drinking it at 1 PPM (roughly the amount added to drinking water in many places, and well below the level in swimming pools) is considered harmless to humans and kills pathogens, so it's a net positive. It's possible that chlorine at 1 PPM causes cancer, but that's a claim that would require evidence.
The same argument applies to pesticide or any other substance. Without talking about specific numbers, it's just speculation.
In terms of fresh meat and vegetables, it's pretty much all grown/produced in Australia. Anything canned / dried is often imported though. Things like rice or coffee beans you technically can buy Australian grown but you'd have to go out of your way to find it.
People don't even know. I had long assumed that it was only the obvious nylon pyramid tea bags that were plastic, and only recently discovered it's _all_ tea bags.
They would have to build the product twice one for mobile chips and one for server, and then there would be functionality discrepancies. Or even worse, the on server one might work better than the on device one that newer phone users get.
If you want hosted AI you can already install the Gemini app or whatever. The only advantage Apple can offer is something that runs on device.
Most of the time this prompt comes up it's actually for a genuine purpose, like spotify trying to find devices on the local network that can play audio, VLC looking for chromecasts, I saw my DJ app ask for local network and discovered it can discover my decks on the network and stream my library over the local network to it.
The problem is this prompt is new so the software doesn't show the user why it's just triggered the prompt and the user has no info to work with.
When a person expands bullet points they add extra information from their own knowledge and research in the process of writing. When AI does it, it adds filler and repetition.
Long form writing itself isn't a problem, it's the empty fake long form we have now.
There's a difference between a long text and an essay. You wouldn't spend hours typing a message and formatting it with headlines for example. You wouldn't insert loads of unessorasy creative writing techniques in to a message asking for help.
Except 99% of the time they are asking it's because they explicitly need a real opinion or the info couldn't be found via LLMs. But instead of giving an "I don't know", they paste back an wall of text with an incorrect answer that the sender hasn't even read or verified to be true.
At least with "I don't know" the asker can move on to someone who might know faster.
It reminds me of how LLM hallucination is attributed to "I don't know" being underrepresented in training data, and it being a better strategy to guess on evaluations rather than admit not knowing.
Different reward function, but the same behaviour emerges.
We'll see that improve as people move onto synthetic training data-- something now possible that we have sufficiently smart LLMs to create enough of it.
The idea is that you generate fake llm transcripts using your classical training data. E.g. look at some training data, generate q/a transcripts. Generate radom questions, RAG against your whole dataset and look for relevant stuff, if there is nothing there, train a "I don't know." reply.
A moderately sized LLM operating some tools to access more information behind the scenes, perform tests and correct its own errors can write transcripts simulating a much larger and smarter llm.
We have all these terrible illnesses that we ascribe to bad luck, and then all of these new chemicals we haven't fully studied yet being sprayed on everything.
reply