Dan Luu
Why didn't Anthropic do effective false positive rejection with their Mythos/Glasswing vuln reports? In danluu.com/ai-coding/#m..., I mentioned a colleague finding that the reports we got were mostly false. That seems common? E.g., gregkh on kernel reports and www.vulncheck.com/blog/anthrop...
Did Ralph loops ever work? They were popular for a while, but when I did a comparison, they underperformed. There were all these theories about why Ralph loops were effective ("invert the latent space of the model"), but AFAICT people just didn't run the comparison? danluu.com/pl-tokens/#r...
Dan Luu reposted
Maybe the best thing on the internet at the moment: * An AI PR pitching platform has confused Kyle Kingsbury, MMA fighter and podcast host with Kyle Kingsbury, software engineer and gay leather man * Good Kyle decided to start a podcast where he interviews confused manosphere influencers
Interesting to see Steve Yegge say he never successfully built anything with Gas Town. I mentioned not finding these super vibed orchestrators useful because the reliability was too low (w.r.t. completing non-trivial tasks). Turns out the author of the most famous one had the exact same issue.
Agentic test processes, LLM benchmarks, and other notes on agentic coding from Galapagos Island: danluu.com/ai-coding/
How do programming languages impact token efficiency and correctness? danluu.com/pl-tokens/
The bulk of www.ftc.gov/system/files... is about standard dark patterns, but there's a fairly good-sized section on the use of 3rd party trackers. It would be really interesting if those end up carrying a real cost the company. Less interesting if this results in another GDPR/EU-banner-like thing.
Exercises in benchmarking and evals, part 7: performance napkin math, DeepSWE / Senior SWE-Bench, and winter tires danluu.com/exercise-7/
Agentic test processes, LLM benchmarks, and other notes on agentic coding from Galapagos Island: danluu.com/ai-coding/
A version of Missile Command for the Commodore 64 where the bottom of your screen is the game state in memory and missiles cause memory corruption: csdb.dk/release/?id=.... In the video below, a missile broke my controls and caused my cursor to get stuck moving down and to the left.
I find it interesting/surprising that Tesla topped the www.iseecars.com/most-dangero... fatalities per mile ranking from 2018-2022. Fatality rate is strongly negatively correlated with price and weight and Teslas are much more expensive and heavier than average.
The commentary I've seen says Teslas are safe so it must be the drivers but, per danluu.com/car-safety/, maybe it's the cars. The most fatal rated manufacturers (Kia/Hyundai, Dodge, Tesla) all did poorly — Kia/Hyundai, Dodge got the lowest rating and there's a strong case Tesla should have as well.
Dan Luu reposted
Replying to @samuel.fm
i can speak with a little authority on this, i work in this field DCs expend water by evaporation, either in cooling towers or evaporative coolers (water becomes vapor, taking latent heat energy with it into the air and leaving cooler liquid water behind) so we expend liquid water resources 1/
On the 2011-2012 FTC antitrust investigation of Google: danluu.com/ftc-google-a...
Dan Luu reposted
What Factors Explain the Nature of Software? tratt.net/laurie/blog/...
We're coming up on a decade since danluu.com/cpu-bugs/. Not only was the prediction valid for a decade, it looks like the next decade will be worse. Today's bug of the day: a significant fraction of high-end consumer CPUs crash if you run CPU-intensive workloads: www.radgametools.com/oodleintel.htm
Diseconomies of scale in fraud, spam, support, and moderation: danluu.com/diseconomies...
Dan Luu reposted
A thoughtful retrospective by @adamwiggins.bsky.social on the business behind Muse, a tool-for-thought/iPad idea-sketching app. I’m sad that the business didn’t work out as planned, but the software still exists and the underlying goals and values remain good. adamwiggins.com/muse-retrosp...
Notes on the accident where a Cruise self-driving car dragged a pedestrian under the vehicle for 20 feet after hitting them: danluu.com/cruise-report/
Why do people post on [bad platform] instead of [good platform]? danluu.com/why-video/
How bad are search results and how do small, independent, search engines compare to Google and Bing? danluu.com/seo-spam/
When designers have free reign to do things how they want, the result is often quite unusable: twitter.com/danluu/statu..., twitter.com/danluu/statu..., etc. And sure, #notalldesigners, but this is a problem that goes back decades across multiple industries, e.g., the story about when
automotive engineers had to fight designers to get headlights that illuminated the road. They were losing until Consumer Reports started testing headlights for function, which created enough bad PR that engineers were able to push for headlights that actually work: danluu.com/why-benchmark/