I'd argue it would be unethical to not do so. I can see where it may lead to false-positives, but in those instances, it's better to be safe than sorry.
A reasonable and responsible approach could be to instruct the child to seek a safe adult around them to discuss any material that may be harmful.
For my own kids, I think I'd prefer it not to instruct the child do do anything in any circumstances, unless they explicitly ask how to do something. In cases of health emergencies, I think it's important for my kids to be able to call 911. Maybe these are decisions we can have in the parental settings, so parents can make that decision.
This. Fines are not "the cost of doing business". It's a tool that when leveraged properly can be used to get out of compliant entities, in compliance.
I do weekly notes instead of daily. My problem with daily notes was that managing them was tedious and felt very duplicative (even when I found ways to automate their creation). It made searching for a particular topic or task more difficult because the unfinished item is in 3-4 daily notes cause it gets buried and I have to search through all of them.
Now that I am using weekly notes, I find myself finding information easier and the work to create the note is a 1/5th of the chore since I only do it Monday mornings.
Usually I see that term applied to Turing tarpits, intentionally obfuscated languages, joke/meme languages, or ones with highly heterodox syntax. Zachtronic languages are really none of these and closer to an assembly (reduced) instruction set architecture. What makes them toy-like languages is that the "machine" you're writing assembly for is rather oversimplified, and is in fact totally fictitious.
I like this, already got face ID built into our devices including PCs these days... tap into that. That curbs the whole issue of sending your identifying information to third party websites.
I think a lot of this has to do with the way MCP is being marketed.
I think the protocol itself should only be used in isolated environments with users that you trust with your data. There doesn't seem to be a "standardized" way to scope/authenticate users to these MCP servers, and that is the missing piece of this implementation puzzle.
I don't think Github MCP is at fault, I think we are just using/implementing the technology incorrectly as an industry as a whole. I still have to pass a bit of non-AI contextual information (IDs, JWT, etc.) to the custom MCP servers I build in order to make it function.
The MCP protocol explicitly says that servers are expected to be run in a trusted environment. There have been some recent updates to the spec that loosen this requirement and add support for various auth schemes, but
Understanding this isn't necessarily my audience, I am not a startup, instead a freelance consultant. I could call it a startup, but that would be stretch, and I am not interested in VC funding.
If you don't value your developers perspective in the sales cycle, you are fundamentally doing something wrong. Developer's don't necessarily hold more product knowledge than anyone else, but they sure as shit can tell you how your company can make an extra 100k on a sale by just adding 10 more lines of code.
Sales are incentivized to sell, developers are incentivized to provide impact. Include them.
A reasonable and responsible approach could be to instruct the child to seek a safe adult around them to discuss any material that may be harmful.