doctoboggan

5 ms·

31.
by doctoboggan 1mo ago
A bit of a meta comment: I love when wikipedia articles are posted to HN, I always click on them as I know they will be interesting. On one of the recent wikipedia article posts (Glacier Mice I believe) someone commented that there should b
32.
by doctoboggan 1mo ago
Yes you are right not sure how I flubbed that one. Made an edit in my comment.
33.
by doctoboggan 1mo ago
If you read the "Manufacture" section you will find this: > Wulf Hein and Kurt Wehrberger conducted an experimental replication with the kinds of stone tools available at the time. Removing the base of the tusk took ten hours.
34.
by doctoboggan 2mo ago
> Ajeya Cotra, one of the other authors on the report, wrote a blog post with her takeaways from this incident. She concludes, “Compared to the reward hacks we know of from just six months ago, this incident feels like it’s more than 50%
35.
by doctoboggan 2mo ago
At this point, I think anthropomorphizing the models gives us better insight into expected behaviors rather than continuing to insist they are just simple probabilistic token generators.
36.
by doctoboggan 2mo ago
> Apple’s most powerful Mac raises the bar for local AI Wow, "Local AI" mentioned in the subheading above the fold - it's really awesome to see Apple leaning into this use case and I think it will definitely pay off for th
37.
by doctoboggan 2mo ago
Sorry, I thought we were talking about Human + LLM agent for software development. I misinterpreted your initial statement.
38.
by doctoboggan 2mo ago
> the days when a human can do anything to help AI there are long past; the human can only drag it down. I disagree, currently a human guiding an AI agent is far better (more productive for less money) than an AI agent with very loose no
39.
by doctoboggan 2mo ago
> Local AI would thwart this vision While I agree Local AI would be a good goal, even the huge open weight models that can't be run locally are already thwarting the US AI lab's duopoly. Other companies can and do serve up thes
40.
by doctoboggan 2mo ago
I often think back to the early GPT days before agents as maybe the last time I will deeply learn a technical subject. I think that was when I actually learned the most as I had to stay completely in the loop. I had recently started a new j
41.
by doctoboggan 2mo ago
> That sounds great until you realize that "hey Claude, read this Jira ticket and implement the feature in this code base" isn't really worth $200K/year. When you factor in overhead and benefits, many companies were r
42.
by doctoboggan 2mo ago
I think that should be illegal. I don’t know how to legislate it in a way that makes sense and actually accomplishes the goal (or if that’s even possible), but I would support someone trying to figure it out.
43.
by doctoboggan 2mo ago
Its a proxy fight - the data centers represent AI in most peoples mind, and any animosity towards a nebulous idea of "AI" can be redirected to these concrete, physical data centers. All that being said, if your actual goal was to
44.
by doctoboggan 2mo ago
As someone who sells actually custom created stuff on Etsy, it’s still definitely the best place for that but the alibaba and temu junk is growing on the platform.
45.
by doctoboggan 2mo ago
When Opus 5 came out I felt myself struggling to follow along and at first was wondering if this is the moment the machine surpassed my ability to follow along and be useful. However over time it does appear it's all an artifact of the
46.
by doctoboggan 2mo ago
> Still, think most people don't realize how good LLMs are at 3D these days, especially Fable 5 Can you explain more? Can I use Fable 5 to generate a 3d model from a couple of photographs?
47.
by doctoboggan 2mo ago
Yes, this really ought to be trained in (or at least RLHF'd in) but that would hurt engagement numbers so the opposite is done instead. These are tools and it would behoove us all to keep that top of mind. Dangerous tools that are not
48.
by doctoboggan 2mo ago
> The most egregious example was when it called my draft of a LinkedIn post “engagement bait”—not saying I disagree, but there’s no need to say it to my face! Maybe it did need to say that to your face. Please don’t post LLM slop to Link
49.
by doctoboggan 3mo ago
Isn't this sort of like asking "Why does getting shot kill some but not other?" It depends on which part of the body was shot, and which part of the DNA was damaged.
50.
by doctoboggan 3mo ago
According to these charts I should switch from Fable to Opus in Claude Code now?
51.
by doctoboggan 3mo ago
git's data model is incredibly powerful and flexible, but its UX is famously... interesting: https://stevelosh.com/blog/2013/04/git-koans/
52.
by doctoboggan 3mo ago
Similar to the story of George Dantzig, who was late to class and solved two open problems in statistics because he mistook them for homework, I think the current batch of frontier LLMs are chained up by knowing which problems are supposed
53.
by doctoboggan 3mo ago
Yeah agreed, from one standpoint I couldn't care less that they did a "distillation attack", but I am interested in knowing if China is able to develop open weight frontier models without the prior existence of a huge model t
54.
by doctoboggan 3mo ago
Does the apple passwords app work this way (log in to a public machine by scanning a QR code?)
55.
by doctoboggan 3mo ago
The legal argument is that we gave this data "voluntarily" to the data broker so the government no longer needs a warrant. This is an area that desperately needs new laws to catch up to the reality of what is going on, but I don&#
56.
by doctoboggan 3mo ago
I have a side business selling custom fingerprint jewelry and I use gemini nano banana to clean up customer submitted fingerprint images. This was a step I used to do by hand at 10 - 15 minutes per image and nano banana is the first model t
57.
by doctoboggan 3mo ago
I feel bad for the engineers at Cursor who have to use Grok in these sorts of experiments.
58.
by doctoboggan 3mo ago
I think the SotA is moving too fast for the production timelines of an ASIC, wouldn't you think? People are just now coming out with LLAMA ASICS but who would want to use LLAMA? Or I guess you are arguing that the models _now_ will be
59.
by doctoboggan 3mo ago
What are the steep angle arrows indicating? Too steep for an escalator/stairs, but not 90 degrees like an elevator would need to be. Anyone know?
60.
by doctoboggan 3mo ago
> Is there anyone credible who thinks this is a plausible pathway for SpaceX to make huge amounts of profit? Scott Manly (who I think is credible) has a video where he goes over the logistics of SpaceX's space based data centers. He

More ›