liamwire

> LLMs give no attribution. I deserve to be credited for my fractional contribution to the collective output of humanity.

I disagree, respectfully. You’re unable to accurately credit the myriad of people, sources, and works that’ve all educated and influenced you along the way. Suppose even if you could, you’d not be able to resolve their dependencies ad infinitum. Nor would it even make sense—arguably, as a species we’re defined by the compounding and shared nature of the knowledge we learn, implicitly and explicitly. Shoulders of giants, and all that.

If your contributions were a significant portion of the entire training corpus, then perhaps we’d have a separate discussion to have, but there’s no single person for which you can reasonably argue that’s true. Maybe for authors of hyper-niche topics or cutting edge research?

sama and OpenAI’s CPO Kevin Weil both suggested this is coming soon, as a direct response to DeepSeek, in an AMA a few hours ago: https://www.reddit.com/r/OpenAI/s/EElFfcU8ZO

Your last point matches what I’ve seen some people (simonw?) say they’re doing currently: using aider to work with two models—one reasoning model as an architect, and one standard LLM as the actual coder. Surprisingly, the results seem pretty good vs. putting everything on one model.

YR24 is estimated to be less than a third the size of Apophis. Side note, we’re only about 4 years away from that asteroid coming within 32,000 km of the planet, closer than the orbiting height of geosynchronous satellites.

If YR24 were to hit the planet, the article puts the impact energy on par with a 10 MT nuclear bomb, or similar to the Tunguska event:

“Everything within three or four kilometers would be incinerated,” Tonry says. “Everything out to maybe 10 kilometers is smashed. […] There would be a huge fireball that would start fires out to 15 kilometers, something like that. It would kill a lot of people if they haven’t moved out of the way.”

So, not nothing, but also not an existential threat even for most individual nation states, and also a 98% chance of being entirely uneventful.

https://en.m.wikipedia.org/wiki/99942_Apophis

https://en.wikipedia.org/wiki/Tunguska_event

Would this extrapolate to the thousands of lightbulb prototypes it took to arrive at the first working one? Rinse repeat for your preferred innovation.

Resource allocation in this context isn’t at all binary.

Huge fan of wrapping simple PowerShell tools with Gum for making things a bit nicer for colleagues to use, Charm’s entire catalogue is art

Great question, I’m not sure about the answer. I’ve heard those that meditate describe similar experiences, and have had some profound perspective shifts trying it myself years ago. I can’t also rule out the age I was at the time, being one where’d you’d expect to go through new and transformative experiences anyway by virtue of still growing and maturing.

But my intuition says no, it really does feel like those were peak, pivotal experiences that still stand out as some of the most significant in my life. Not to say it’s not possible, but maybe more so that in my little corner of the world with the relatively limited experiences available to me at the time, I’m not sure I wouldn’t have simply continued to tread the very uninspired life path I was on.

In a way it felt like waking up in the middle of a dream, and realising I could go back to sleep, or get up and change my circumstances. Probably a bit of a cringe analogy, but it feels about right — it was still work and a conscious choice to make positive changes, but prior to then it hadn’t even occurred to me that I could. It’s not lost on me that for positive stories like mine, there’s many people that could counter with negatives.

Drugs aren’t safe, but neither is a life unchallenged I think. In my case that was what I needed I suspect, a challenge to my own views on myself, other people, the world, and what’s possible. Therapy might’ve gotten me there too, for all it’s worth, but I don’t know that I would’ve considered it prior.

My take on this is that, despite an ever-increasingly connected world, you still need an assistant like this to remain available at all times your device is. If I can’t rely on it when my signal is weak, or the network/service is down/saturated, its way of working itself into people’s core routines is minimal. So either the model runs locally, in which case I’d argue OpenAI have no moat, or they uncover some secret sauce they’re able to keep contained to their research labs and data centres that’s simply that much better than the rest, in perpetuity, and is so good people are willing to undergo the massive switching costs and tolerate the situations in which the service they’ve come to be so dependent on isn’t available to them. Let’s also not discount the fact that Apple are one of the largest manufacturers globally of smartphones, and that getting up to speed in the myriad industries required to compete with them, even when contracting out much of that work, is hard.

Genuinely thought we’d never see the day. My feelings on Ulbricht are mixed and have evolved over the decade he’s been in prison.

However, the Silk Road allowed me to try LSD as an 18 year old in a safe(r) way than those that came before me.* It was those experiences that revealed I’d been depressed most of my life, and that it also didn’t have to be that way, by way of experiencing what that would feel like. I went on to seek new experiences, make new friends for the first time in my life, engage with professional mental health support, went to university, and started multiple businesses. It also introduced my staunchly-atheist self to the experience of spiritual/transcendental experiences, and how those can exist separately from, and don’t require, belief in deities or religion.

It can’t be said where I’d have wound up without those experiences, but my own understanding of myself feels pivotally tied to something I couldn’t have gone through without Ross’ actions. Still, I acknowledge it appears more likely that not he tried to have people killed, and regardless of the circumstances surrounding this, that is condemnable.

*Had it not been for an anonymous group at the time, The LSD Avengers, posting reviews using gas chromatography mass-spectrometry and reagent tests of suppliers on the site, I wouldn’t have had the confidence to take the risk of trying what I’d received. LSD is physiologically safe, not to say anything of any psychological risks, but knowing the dose allowed me to enter into the shallow end of the pool, so to speak. Common substitutes however cannot have the same said of them.

If I’d lived in a time and place that allowed for state-funded drug testing (something my own state has in fact recently abolished despite wildly successful trials), perhaps things would’ve not required a Ross Ulbricht to exist in my case, but I see this as a failure of the system and of drug prohibition as a whole.

Ross would’ve existed one way or another I believe, for better or worse, by another name, had he chosen another path. Now he gets the chance to try his life again. I felt the same way.

No luck catching them electrons, then? It’s just the one electron, actually.

https://en.wikipedia.org/wiki/One-electron_universe

Not to speak for him, but I’d suspect Simon would be using a terminal emulator. He often mentions preferring to use his own command line tool, named LLM, as it allows one to log everything to an SQLite database for later exploration.

I found his book Waking Up to be a phenomenal resource in discussing meditation, psychedelics, and spirituality without reaching for religion/unscientific guesswork. Free Will and On Lying were both interesting, shorter reads that I thought to be very persuasive as well.

Qubes is far from esoteric in spaces where security is paramount. Your lack of familiarity with it doesn’t mean it’s obscure. It’s more a tool for a specific subset of people and purposes, rather than an OS meant for wide adoption.

Highly recommend the experience, it was darkness unlike anything I’d ever seen before. Excuse the phrasing.

You truly believe we’re on a timeline that involves the replacement of anaesthesiologists, emergency medicine physicians, trauma surgeons, and so on, within a 20 year timeframe? AI progress in the last few years has been astounding, but the gaps between where we are and a true all-human-labour-is-inferior scenario is almost unfathomable.

You seem quite emotional about this. Is there something more to your argument against this other than your feelings about how others address their grievances in an unequal power structure?

Sounds entirely at odds with any accessibility requirements.

Absurd, what makes you think a criminal conspiracy of the scale you’re speculating about could possible stay secret, long term?

Ludicrous. Extraordinary claims require extraordinary evidence, etc.

Can you speak to some of the harder or more interesting challenges you faced during that time?

I have a hard time believing the ‘since’ part of that description. Intuition suggests the latest iPhone would take that crown each year.

Delighted to see this here — the Kurilpa Bridge is less than 50 meters from my front door and is something to behold each time I get the chance to use it.

You make a good point, though I don’t think it’s counter to mine.

A poor strawman. I emphatically wish the best for the Iranian people, but only their government would have you conflate the two.

That’s a lot of words to argue agent the right for humans to self-determine their own governance and fate. Sincerely, I see no compelling argument that can be made in support of the restrictions Iran places upon its citizens. Please make one, with specifics, if you’re interested in discussing it.

I’m aware of the history of Iran, and your comment is entirely irrelevant. I can’t celebrate the advances of a country that’s openly hostile towards fundamental human rights. Nor do I think anyone else should, either.

It’s awesome to see totalitarian theocracies make advances in technologies that have parallel military uses?

Unless this is specifically relating to API access, I don’t think it’s correct. I’ve been paying for ChatGPT via the App Store IAP for around a year or less, and I’ve already got both o1-preview and o1-mini available in-app.

I suspect it’s far lower than that.

Beautiful game that’s very quick to pick up, well done.