Profile avatar
scottleibrand.bsky.social
Built Open Artificial Pancreas (OpenAPS) with Dana Lewis. Day job at Netskope. DMs are open or email [email protected].
168 posts 334 followers 76 following
Regular Contributor
Active Commenter
comment in response to post
I loved it! simonwillison.net/2025/Jun/10/...
comment in response to post
I suspect (on priors) 1 and 3, to fund buyouts for 4.
comment in response to post
Good reminder, thx. Can it do all the same tool calls from there as o3 does in ChatGPT?
comment in response to post
Report it to the marketplace (e.g., Amazon “Unwanted Package” form) and optionally ReportFraud.ftc.gov Make sure none of your accounts were compromised (if not, likely they made a new one shipping to your address). Keep/sell/donate the stuff.
comment in response to post
Asking that of o3 gives a pretty good overview: chatgpt.com/share/68394e...
comment in response to post
Block 1 worked. If they have to reduce capability to get reliability they will. Would mean more tanker launches per flight and raise costs.
comment in response to post
...But too many critics of those stasist ideas try to shove the underlying problems under the rug. With this post, I"m trying to help us hold both things at once. Read the full post on AI Frontiers: www.ai-frontiers.org/articles/wer... Or my substack: helentoner.substack.com/p/dynamism-v...
comment in response to post
The map always says Closed Winters, no matter the season. Sounds like it’s refusing to route people that way, and people are noticing the more visible Closed Winters designation. Hope they get the (invisible) routing restriction fixed quickly.
comment in response to post
We have enough evidence to suggest doctors should probably be using AI themselves to get a second opinion on their diagnoses. They can choose whether to use that advice, but not using it seems increasingly like voluntarily ignoring an important tool that can help patients.
comment in response to post
Must be as in “should be”? Or must as in “surely is”? There is a lot of land out there. Good locations are worth using for multiple things, but it doesn’t always make economic sense for more rural areas…
comment in response to post
Maybe that’s part of why they’re so high, so you can just drive under them for repairs?
comment in response to post
So if those numbers reflect on-the-ground reality (which seems a bit implausible to me, but IANAE) wouldn’t Q4 have been the latest record and Q1 not a record quarter?
comment in response to post
And if that much utility solar is being reported annually instead of quarterly, the 60% rooftop statistic would be an artifact of that as well, and utility would likely still be a majority of actually installed capacity?
comment in response to post
I’m confused. Total installations were way down from Q4? Was the first sentence also supposed to say rooftop? Why the big Q4 jumps in reported utility solar the last two years?
comment in response to post
Even its authors were skeptical. Good overviews from o3 and Grok agree: chatgpt.com/share/682b90... grok.com/share/bGVnYW...
comment in response to post
The latter: chatgpt.com/share/6823ef... “If hotspot speed is fine on your Pixel but slow on the iPhone, your plan almost certainly applies a hotspot‑specific throttle that iOS dutifully exposes. Pixels don’t volunteer that they’re tethering, so the carrier grades them as normal phone traffic.“
comment in response to post
I’d be interested to help fund scaling this up if they/someone can get it off the ground. If we can validate the ROI remains <1y as it scales up, this could be several times more effective than cash (e.g. via GiveDirectly).
comment in response to post
Does the Pixel allow carriers to enforce the terms of their plan like the iPhone does? Or does it make hotspot data appear identical to the carrier to on-phone data?
comment in response to post
Make sure you’re on a carrier plan that supports full speed hotspot. Some of them rate limit it.
comment in response to post
IANAS, but FWIW o3 and Grok 3 agree to a yes on both counts. (And TIL what the AMS even is, so my vote definitely doesn’t count.)
comment in response to post
The AI futures project has a post on why America wins the AI race, and with it the future, and the short version is America has more compute than China, and will keep that compute lead long enough into the intelligence explosion (#29): blog.ai-futures.org/p/why-americ...
comment in response to post
"If you want to prompt ChatGPT 40 times, you can just stop your shower 1 second early." "If I choose not to take a flight to Europe, I save 3,500,000 ChatGPT searches. this is like stopping more than 7 people from searching ChatGPT for their entire lives."
comment in response to post
Got the Spanish grid operator data on generation by source type working now too. Confirms the ENTSOE data that it remains solar/wind/hydro that is still operating. Nuclear/gas/coal remain basically offline for now - presumably tripped off when demand fell (or vice versa depending on the root cause).