I've personally liked 4.6 the best to date, preferring it by far to 4.7 and 4.8 (even with these on max effort!), both in Claude Code and for non-coding tasks in the chat UIs.
Still early but from my first few interactions with Fable on high in both settings, it feels like it might finally dethrone 4.6 for me, but time will tell.
Hoping it doesn't get nerfed and eventually comes back to the subscriptions.
Well, it is arguable that it is "laughable" that there is a strong push to try and regulate/contain these systems to avoid human extinction when they can't tell you the number of weekdays in a month (and fail in this manner).
I'm no AI skeptic (and am aware the above does not seem to apply to GPT4), but the general point stands - finding these limitations is an interesting endeavor, and when they're particularly severe or unexpected, newsworthy.
Uhm - maybe train a secondary NN that scores summaries on their factual accurateness/quality? Anything under a given threshold is either sent for manual review or re-ran through the LLM until it passes.
One thing I noted was that the median price paid by foreign buyers was not as high as I expected when compared to the median price for all buyers (280.6k compared to 259.6k).
This indicates to me that these foreign buyers are not necessarily institutional investors or companies, but foreign citizens, so the segment of the market they're in is probably not the multi-million property segment, but the more "standard" residential properties market a normal US resident would be in. The article does somewhat confirm this:
>"Foreign buyers include those living in the U.S. and overseas, but the majority (60%) were recent immigrants and foreigners who live in the U.S. for work, school or other reasons."
If you also take into account that there is an effect of the purchases made by these buyers on the "median price for all buyers" cited above (and so the real difference in median prices between foreign/domestic buyers is higher than what is suggested by those figures), I think the effect is somewhat diminished (as this pushes them into a more "upscale" market).
Even with the above, I still think it's relatively safe to assume this will have a downwards impact on US residential prices, at least in these states (but I'd need precise figures as to what share of homebuyers are foreign).
*edited the last paragraphs to correct something I had gotten mixed-up on
Still early but from my first few interactions with Fable on high in both settings, it feels like it might finally dethrone 4.6 for me, but time will tell.
Hoping it doesn't get nerfed and eventually comes back to the subscriptions.