HNHacker News
TopNewBestAskShowJobs

rogerkirkness

1,440 karma · joined July 3, 2018

Personal: https://rogerkirkness.com

Work: https://nonducorduco.com

submissionscomments
rogerkirkness··on iCloud Mail or Gmail?
I connected Gmail to the Mail app on MacOS via IMAP, backed that up to iCloud, and setup my personal custom domain with iCloud Customer Domains. I then forwarded the Gmail to it and deleted all the mail on the Gmail side. I would not go back to Gmail, I considered getting rid of the Google account entirely recently. Hide My Email is also really nice and easy if you use Safari.
rogerkirkness··on Trump says U.S. will not approve solar or wind power projects
“The days of stupidity are over in the USA!!!” - Trump in relation to this announcement. Lol.
rogerkirkness··on Shamelessness as a strategy (2019)
A lot of modern movies are metamodern. The example on the Wikipedia is Bo Burnham's INSIDE, which is very metamodern. But I'd argue most movies are at least somewhat at this point self aware, fourth wall breaking commentary. Top Gun is an example of a modern movie - not metamodern, where there is diversity but it's not at all self aware, fourth wall breaking or commentary.
rogerkirkness··on 'Work-Life Balance' Will Keep You Mediocre
Yeah I agree, there is a total absence of values from silicon valley at this point. It is no different from 1980s wall street basically.
rogerkirkness··on Google is watching
Highly recommend iCloud Private Relay plus Safari. Sites often think I'm in New England, or Montreal, or a bunch of places I'm not, seemingly at random.
rogerkirkness··on Shamelessness as a strategy (2019)
It's hard to unsee metamodern in movies once you see it. I did a 'self aware waldo' monologue in high school drama class, didn't realize it was meta modern.
rogerkirkness··on 95% of generative AI pilots at companies are failing – MIT report
Is the purpose of gardening to be arms and legs for ChatGPT to grow a garden?
rogerkirkness··on Claude Opus 4 and 4.1 can now end a rare subset of conversations
It seems like Anthropic is increasingly confused that these non deterministic magic 8 balls are actually intelligent entities.

The biggest enemy of AI safety may end up being deeply confused AI safety researchers...

rogerkirkness··on China's Lead in Open-Source AI Jolts Washington and Silicon Valley
Commoditize your compliment baby. China leads in most frontier technologies, might as well just commoditize the one they have been stifled from competing on.
rogerkirkness··on Silicon Valley's Growing Obsession with Having Smarter Babies
Silicon Valley in some sense has no values, it is enlightenment liberalism run wild. Some thing about that are fascinating and good, but others not so much.
rogerkirkness··on Why are there so many rationalist cults?
Because they have serious emotional maturity issues leading to lobotomizing their normal human emotional side of their identity and experience of life.
rogerkirkness··on AI agents fail tasks 70% of the time
Agents went from 10% to 30% reliable this year, which is still a big deal.
rogerkirkness··on Ask HN: Could we pool our machines to run more powerful AI models locally?
The meta project for SETI@Home is called BOINC: https://boinc.berkeley.edu but I'm not sure if it focuses on running large open weights models for inference as much as compute in general. I agree with the premise though.
rogerkirkness··on Linear sent me down a local-first rabbit hole
Yeah I built my first startup on Meteor, and the prototype for my second one, but there was so many weird state bugs after it got more complicated that we had to eventually switch back to normal patterns to scale it.
rogerkirkness··on Linear sent me down a local-first rabbit hole
Reminds me of Meteor back in the day.
rogerkirkness··on 'I Feel Like I'm Going Crazy': ChatGPT Fuels Delusional Spirals
I do think a major part of AI is godless liberals trying to find containment for the uncertainties of life. And I mean that in a neutral way as a godless liberal.
rogerkirkness··on How AI conquered the US economy: A visual FAQ
Winner takes most is now true at the global economy level.
rogerkirkness··on Stop saying AI makes us work longer
The whole idea is for people to work on things that are interesting and valuable to customers at the same time. AI is for things that are valuable to customers but not interesting to work on.
rogerkirkness··on OpenAI's "Study Mode" and the risks of flattery
I would research teleological thinking, some people's brains have larger regions associated with teleological thinking than others.
rogerkirkness··on 'Superintelligence' Will Create a New Era of Empowerment, Mark Zuckerberg Says
Question: if you give everyone way more power, what is the outcome?
rogerkirkness··on Sleep all comes down to the mitochondria
Yes, assuming you get exactly the right amount and somehow also never too much of the bad UV light. It's a finicky thing to get right, but bottom line yes.
rogerkirkness··on Sleep all comes down to the mitochondria
Highly recommend red light therapy for this. There's a spreadsheet that contains [1] all the scientific research does on effect on mitochondria.

[1]: https://docs.google.com/spreadsheets/u/1/d/1ZKl5Me4XwPj4YgJC...

rogerkirkness··on Web fingerprinting is worse than I thought (2023)
Yes by a lot.
rogerkirkness··on Silicon Valley AI Startups Are Embracing China's Controversial 996 Work Schedule
Any senior engineers facing this feel free to reach out, we are now four day work week at full comp (serious). Remote, not 'Silicon Valley'.
rogerkirkness··on Reflections on OpenAI
Calvin is the founder/CTO of Segment, not old but also not some doe eyed new grad.
rogerkirkness··on The End of SaaS?
Gitbook and Webflow replaced with our own homegrown static sites on VitePress.
rogerkirkness··on Spending Too Much Money on a Coding Agent
Early stage founder here. You have no idea how worth it $200/month is as a multiple on what compensation is required to fund good engineers. Absolutely the highest ROI thing I have done in the life of the company so far.
rogerkirkness··on The End of SaaS?
We replaced two minor SaaS vendors this week with vibe coded alternatives, so yes I do think that it's real. Our core systems like HRIS, ERP, Github, email, etc. no and no plans to replace or foresee being able to replace those ones.
rogerkirkness··on Reddit in talks to embrace Sam Altman's iris-scanning Orb to verify users
Not so good for the business of harvesting human interactions for AI training. Need to verify it's certified 100% real human interactions.
rogerkirkness··on Ask HN: What weird ways are you using Claude Code?
We moved both our marketing website and documentation website into markdown in our monorepo. Now when we are working with Claude, we actually have it update our marketing (reviewed by product marketing in PR workflows) and our documentation (reviewed by support engineering) right inline with the code changes themselves. In a way, it allows us to change our entire customer facing experience end to end in a single change process via PRs, compared to before where those would be downstream things we would do.

I'm now thinking about whether to move internal operations things like company policies, etc. into the codebase and/or data and somehow be able to use Claude to operate across them via Claude Code, for example when we change a policy or when we change a strategy definition like ideal customer profile, it could also go and update that across all the places it is referenced in our knowledge base and then allow me to update that via PR the way we do with code now. Really you could do government policy this way too, change a constant value and have it effect all the written policy

← PreviousPage 4 of 32Next →