As someone that's been working on a game with Claude Code in a more human-in-the-loop, iterative fashion, I have to say that OP's claim that "LLMs barely know GDScript" does not match my current experience at all even though it seems to have matched yours. Maybe it was true a while ago in both cases; how long ago was your "vibecode" attempt? I've gotten completely fine GDScript and even decent perfectly functional if placeholder-quality TSCN files from Opus 4.5, 4.6, and Sonnet 4.6 with very little issue and no special tricks; just a CLAUDE.md file, the project itself, and going through plan mode before each change. I did start from a playable project with a fair amount of hand-written scaffolding already in place, and I have no idea if that would make a difference. Every once in a while there will be some confusion that I get something appropriate for Godot 3 instead of Godot 4, but never Python despite the similarities of the language.
I agree, Claude does a great job of synthesizing Godot docs, and even of writing unit and integration tests with GUT. I have not tried any e2e testing but I’m not convinced that you couldn’t get a good result there, too, depending on the kind of game.
They did not, you get the same date range and the same graph shape going to FRED and pressing the "1Y" option, and the series includes the first two months of 2026 so it's 12 months: https://fred.stlouisfed.org/graph/?g=1SGzm
However, the chart settings were actually modified to hide/deemphasize the earlier decline: the the index date was changed. 2025-02-20=100 in their graph, default of 2020-02-01=100 would have the chart start at 64 and rise to 71.44.
It unfortunately didn't get very much attention here at the time (2021), but "Sigmoids behaving badly: why they usually cannot predict the future as well as they seem to promise" at https://arxiv.org/abs/2109.08065 is related.
I won't bother defending Google-style personalization as it exists for their search results, but since collisions in terminology across fields are common, it's not that hard to see how actual, thoughtful personalization could be useful. Someone searching for "Kafka" is going to want very different results based on whether they're thinking of software or literature. Opinions may also differ over the usefulness of sources, even for people ultimately interested primarily in facts; I find Kagi-style personalization (make your own domain list) very useful, but across Kagi's userbase Reddit is simultaneously one of the most lowered, most raised, and most pinned domains: https://kagi.com/stats?stat=leaderboard
Anecdotally I find myself appending 'reddit' to search terms very frequently. It's effectively shorthand for "I want to read about peoples direct experience with this thing", and reddit is huge and well crawled by search engines. It's astroturfed to hell especially around political topics, but I feel like it's easy to tell when discussions about random products are authentic.
This could be "bad, actually" if it gives an incorrect impression that power tools are unequivocally safe, rather than somewhat risky but usually safe when used correctly.
You're right, but one presumably would still teach kids to treat this tool with respect. And given that, it seems safer to me as this won't hurt them when they get careless (as kids are wont to do). That way you get a chance to reinforce the safety lesson before they graduate to the dangerous stuff.
I'm finding that a lot of parenting is teaching my kid that safety is something you have to do, and risks are something you have to look for and understand. For example, brushing your teeth is usually safe, but you shouldn't brush your teeth at a dead sprint down the stairs.
Not sure why you've been downvoted so heavily. That seems like a misuse of the downvote purpose.
But yes, I kind of agree with other commenters here in that maybe teaching absolute respect of a knife/table saw/power tool and its power to maim is a really important lesson that this sidesteps?
What exactly is so terrible as long as you're willing to take a small to moderate risk of getting a PVC pipe stuck in the ground and keep in mind the presented cautions -- consider the water irrigation-only/non-potable until tested (possibly even if tested, although that's not what the site says)?
While the abstract does say there's "rising evidence that leprosy has become endemic in the southeastern United States", the actual title is "Case Report of Leprosy in Central Florida, USA, 2022" and that's a more honest description of what this is.
The second reference -- https://www.hrsa.gov/hansens-disease -- shows cases in the entire US bouncing from around 160 to 220 since _2011_ up through 2020: "Most (95%) of the human population is not susceptible to infection... Treatment with standard antibiotic drugs is very effective." The Florida dashboard which is the third reference shows 14 cases in that state total in 2021.
reply