clbrmbr

8 ms·

121.
by clbrmbr 10mo ago
The tearing was unexpectedly disturbing! Suggestion: use an accelerometer data on mobile and use that to directly replace gravity. I expect to be able to tip the phone to drape the cloth, and shake the phone to get waves of motion.
122.
by clbrmbr 10mo ago
Rust version?
123.
by clbrmbr 10mo ago
Awesome blog! Looking at the code I feel like there’s a kindred soul behind that keyboard, but there’s no About page afaict. Who beeth this mysterious writer?
124.
by clbrmbr 10mo ago
Ofc a savings account has risk in real terms. But I assume GP was referring to risk in terms of losing principle in dollars. There’s still some risk short of a global financial collapse where the FDIC rules are weakened, perhaps by making t
125.
by clbrmbr 10mo ago
When I walked across the crater as a kid, I remember there was an inner crater that I was told had filled up with lava back in the 80s and then drained down leaving a deep well. Does someone have a map of the historical eruption locations w
126.
by clbrmbr 11mo ago
Huge thanks to the mods and YC for creating this space. HN is legendary in its own time. Hoping @prodigycorp and the result of us can enjoy another 15 years of thoughtful hackish conversation and news.
127.
by clbrmbr 11mo ago
Do you plan to operate in USA (SEC rules)? The investment structures could use description in English. Does investing in these instruments make the investor a co-owner of the business? (Like a member of an LLC)? Or is it a revenue sharing c
128.
by clbrmbr 11mo ago
Does anyone have a benchmark that clearly distinguishes the larger models? I would think that the high parameter count models would have capabilities distinct from the smaller ones, that would easily be read out. For example, Opus 4 has app
129.
by clbrmbr 11mo ago
What do you mean, they typically get half the assets and a sizable chunk of the other partner’s salary in alimony that they don’t need to give up if they do become employed, and then if childcare is needed typically this would be an extra c
130.
by clbrmbr 11mo ago
Same issue here for me. Downdetector [1] agrees, and github status page was just updated now.
131.
by clbrmbr 11mo ago
The folks who run the public NTP pool really ought not to make it easier to pay them money to use it commercially. I submitted a request for commercial use via their online form but never received a response.
132.
by clbrmbr 11mo ago
But my time is valuable… but I would pay OpenAI's V2V rates for the privilege of stringing along scammers. I need a button “transfer to time waster agent”.
133.
by clbrmbr 11mo ago
I’m one of the people sending these screenshots, and glad to receive them. It helps jog my memory to see a screenshot from the original context, with syntax highlighting or especially log output from our besoke logging system for firmware.
134.
by clbrmbr 11mo ago
I love making nettle tea over a campfire. Drink the tea, then eat the nettles. It tastes very nourishing. Bagged nettle tea doesn’t come close. (Northeastern USA.)
135.
by clbrmbr 1y ago
Now it’s The 12 Days of Code.
136.
by clbrmbr 1y ago
Do we have any institutions to point to, other than the Roman Catholic Church, that have survived this long?
137.
by clbrmbr 1y ago
How can this be? I had great luck with GPT3 way back when… and I didn’t have function calling or chat… had to parse the JSON myself, extraction “action” and “response-text” fields… How has this been so hard for AMZN? Is it a matter of token
138.
by clbrmbr 1y ago
OpenAI goes into great detail on hallucination rates of GPT5 models versus o3 in the GPT5 System Card [1], section 3.7. [1] https://cdn.openai.com/gpt-5-system-card.pdf#page12
139.
by clbrmbr 1y ago
Claude Koan
140.
by clbrmbr 1y ago
Customer Service Lead | Bond Home | ONSITE Cresskill, NJ (15min from NYC) Bond Home [1] Co-founder & CTO here. We're hiring a Customer Service Lead to elevate our support team with cutting-edge AI tools like Claude Agent SDK and re
141.
by clbrmbr 1y ago
Thanks for the correction.
142.
by clbrmbr 1y ago
As soon as you have more than one shareholder, unless it’s in your charter to support OSS, you need some justification. Corporations only “want” to make a profit unless they have another explicit mission at the highest level of governance.
143.
by clbrmbr 1y ago
But the there’s also the negative psychological impact on the user having the model so strongly agree with them all the time. —— I cannot be the only one who half expects humans to say this to me all the time now?
144.
by clbrmbr 1y ago
Unclear how this can be the case in general. Opus is apparently a much bigger model.
145.
by clbrmbr 1y ago
Indeed. I’m having success using it as a tool for requirements querying. (When a sales person asks “does product A have feature X” I can just ask Claude because I’ve got all the requirements in markdown files.
146.
by clbrmbr 1y ago
It depends on the domain. There are a lot of critical utilities in the systems space maintained by volunteers. The “xz” compression library was one recent infamous example where an exhausted volunteer maintainer was social engineered into a
147.
by clbrmbr 1y ago
I’m not an analyst but a product owner and developer. Sometimes I feel similarly. However, I notice a very interesting thing when I turn the AI off for a day and go back to the pre-copilot days: my work is more focused, concise, comprehensi
148.
by clbrmbr 1y ago
Terrible choice of name. DeepSeek developed a historically important model called “R-Zero” (this was the predecessor to R1 that was training without any coldstart SFT, and was very strong but difficult to read chain of thought because it co
149.
by clbrmbr 1y ago
Re: updates: I was just thinking of waiting a few weeks on the updates to allow compromised packages to be discovered.
150.
by clbrmbr 1y ago
Huh. I read TFA in detail (and shared with my team), but I didn’t see any analysis. (?)

More ›