The Contradictions of Sam Altman • Hacker News Board

Re: The Contradictions of Sam Altman

#531

Post by andsoitis » Sun, Apr 02, 2023, 2:31 PM UTC

matthewdgreen wrote: post #524

andsoitis wrote: post #378

Earlier quoted context omitted.

> existing research tech out into the open world (with improvements, of course. That’s quite the understatement, isn’t it? Neither Google nor Meta have been able to demonstrate something to the public that resonates even close the the altitude that OpenAI has.

I'm not sure if this is research quality or product tuning. I tried using Bing's chatbot and I found the experience to be vastly worse than OpenAI's version, yet Bing is ostensibly running on exactly the same platform (GPT-4). I think OpenAI turned the dial to "loquacious and friendly" and we humans are interpreting some pretty-similar underlying tech as having vastly different underlying quality, because we're impre…

The user interface to a tool matters.

Re: The Contradictions of Sam Altman

#532

Post by don_esteban » Sun, Apr 02, 2023, 3:02 PM UTC

logicchains wrote: post #464

piloto_ciego wrote: post #190

Earlier quoted context omitted.

I don’t know that this is true. At very least they both share immense responsibility for causing individual harm. Sure the government may start a war, but that war can’t happen without bombs and bullets, and in America at least those factories aren’t run by the government. There is an intermediate step oftentimes, but I don’t think that necessarily disconnects companies from responsibility. If you work at a guided bo…

>I don’t know that this is true. In the 20th century governments slaughtered wholesale around 100 million of their own citizens (China, Russia, Cambodia), let alone those of other countries they killed in war. There's no measure by which the market comes anywhere near that amount of murder.

The market (german industrialists in case of Hitler, military-industrial complex in case of USA, East India Company in case of GB if you want to go deeper in history) has its quite significant share. Ignoring that is being willfully ideologically blind.

Re: The Contradictions of Sam Altman

#533

Post by artificialLimbs » Sun, Apr 02, 2023, 3:02 PM UTC

DubiousPusher wrote: post #169

Stop asking the market to look after collective interest. This is the job of the government. The ultimate effect of virtue wanking CEOs and corporate governance is to deceive people into thinking democracy is something that can be achieved by for profit organizations and that they can forsake the formal binding of collective interest through law. It's nice if people are nice but it is not a bulwark of the collective…

>> The higher that niceness exists in the social order, the greater its contemporary benefit

Could also be said that the more niceness, the more likely that not-nice actors will violently overthrow the nice ones.

As an example, we have had numerous bee hives at our place. The bees that never mess with us when we come near their hives are more likely to get robbed by another hive and destroyed. The bees that would sting us when we came near were rock solid - they never got robbed or collapsed.

Re: The Contradictions of Sam Altman

#534

Post by arnavk » Sun, Apr 02, 2023, 3:12 PM UTC

DubiousPusher wrote: post #169

Stop asking the market to look after collective interest. This is the job of the government. The ultimate effect of virtue wanking CEOs and corporate governance is to deceive people into thinking democracy is something that can be achieved by for profit organizations and that they can forsake the formal binding of collective interest through law. It's nice if people are nice but it is not a bulwark of the collective…

It's Gandhi not Ghandi

Re: The Contradictions of Sam Altman

#535

Post by noduerme » Sun, Apr 02, 2023, 3:59 PM UTC

IIAOPSW wrote: post #516

noduerme wrote: post #492

To all the detractors of work ethic and the supporters of UBI: What freedom will exist in a world where everyone is reliant on their government for survival, and all mass communication is mediated by or written by opaque AIs? In trading the yoke of labor for free time we would also trade our independence, and any hope of changing the structure of power that had been set in place. I'm so sick of our new style of wanna…

Hard work is only seen as a virtue and sloth a vice because of the context of the world we are currently stuck in. You assume a system of value which might not make sense in the future.

On the contrary, it's the natural state of all thriving organisms to do some form of constructive work. It has also been the human condition and the condition for prosperity as long as we've existed. That isn't to say that leisure and play and pleasure have no value. Not at all! But it's axiomatic that nothing has value without its opposite. And if you look at the wealthy who do not work (as opposed to the ones who do), it's clear that nothing is valued by people without the sense that they've earned it.

Re: The Contradictions of Sam Altman

#536

Post by gdy » Sun, Apr 02, 2023, 4:53 PM UTC

hluska wrote: post #466

whitemary wrote: post #456

Earlier quoted context omitted.

Karl Marx is not only the most influential economist of all time; he is many orders of magnitude more influential than Keynes. Only an American could be clueless enough to think this is up for discussion.

We don’t need this kind of garbage here.

Who are 'we'?

Re: The Contradictions of Sam Altman

#537

Post by cardosof » Sun, Apr 02, 2023, 4:55 PM UTC

jstummbillig wrote: post #478

PostOnce wrote: post #337

Earlier quoted context omitted.

That's all well and good, but it's not well and good to change what you say because the calculus of what will personally benefit you has changed. It's not the same as "new information coming to light", it is instead greed. We've already seen a charity robbed to no end other than Profit. Commercializing OpenAI was for one reason: to enrich the people commercializing it. Not to benefit society, and with nary a thought…

I am astonished how anyone could just make a statement like this, here. It's so sticky, it's so damning and brash. It offers absolutely no proof, while contradicting what sama says themselves. And we are entirely fine with that. Top of the comments after 7 hours. Note, I am not claiming it could not be true. That is entirely beside my point. The issue is, that it could be entirely untrue. And I really mean, every sin…

Burning witches? Call me a radical but I can't help but notice the name of the company is OpenAI, but it's not open at all!

Re: The Contradictions of Sam Altman

#538

Post by davidatbu » Sun, Apr 02, 2023, 5:03 PM UTC

carom wrote: post #530

davidatbu wrote: post #465

Earlier quoted context omitted.

> OpenAI is a company that can barely ship a functioning website, or have proper security for basic stuff. You'd not expect this company to be the future of AI, do you? In case you didn't know: ML researchers are infamous for being bad at traditional software engineering. That's why, I personally don't think whether a company can "proper security for basic stuff" has much correlation with probability of achieving AGI…

This is assuming AGI will be emergent of something that can run in a Jupyter Notebook and not a well engineered system.

Yes, I do think that the most promising path we are aware of to AGI is exactly the route that OpenAI/Deepmind/Anthropic/FAIR/etc are taking.

In fact, I don't think anyone would disagree with that statement.

Re: The Contradictions of Sam Altman

#539

Post by matthewdgreen » Sun, Apr 02, 2023, 6:13 PM UTC

andsoitis wrote: post #531

matthewdgreen wrote: post #524

Earlier quoted context omitted.

I'm not sure if this is research quality or product tuning. I tried using Bing's chatbot and I found the experience to be vastly worse than OpenAI's version, yet Bing is ostensibly running on exactly the same platform (GPT-4). I think OpenAI turned the dial to "loquacious and friendly" and we humans are interpreting some pretty-similar underlying tech as having vastly different underlying quality, because we're impre…

The user interface to a tool matters.

Once you're down to user interface, you're no longer in "insurmountable technical advantage" territory.

But more critically, chat.openai.com is more of a tech demo than a serious product. So one can optimize the UX for "chatty and engaging" without being too concerned for accuracy. Bing, on the other hand, is intended to be a search product. It's noteworthy that adjusting the same model to provide a more accurate search experience seems to have snuffed out a lot of the magic.

Re: The Contradictions of Sam Altman

#540

Post by davnicwil » Sun, Apr 02, 2023, 7:55 PM UTC

runarberg wrote: post #409

davnicwil wrote: post #331

Earlier quoted context omitted.

I think you're agreeing while arguing against my point :-) Like what is collective action, really? It's as you say - a series of individuals interacting, setting a course in a feedback loop, and then more people getting on board with that. It doesn't just happen, it takes individual instigators. It also doesn't just self-propagate. You need individuals, typically the same individuals, to show up regularly and consist…

What you are arguing for is a philosophical position called atomism or reductionism, (which contrasts with holism ). It is a rather old school philosophy honestly (with holism also being old school, but not quite as much) as we are are learning more how important interactions are really to study anything honestly. Modern philosophy of science kind of rejects the notion that you can study anything really by only looki…

Is one difference in the analogy that individual humans have independent agency and do concrete things in the world, where particles don't have agency, and don't really 'do' actions in a sense that is meaningful outside of the system they're in?

I guess by questioning the analogy I get back to my point. Things don't happen in history because of (truly) random behaviours converging on some emergent effect, like in a system of particles. They happen because specific unique (wrt the system) individuals make decidedly non-random decisons to affect reality on purpose (even if cause and effect are not that predictable it still holds that the actions are purposeful and do affect reality) in some way.