partial_accumen

I read this story this morning and have been thinking back to it all day. This wasn’t just some idiot that was too stupid or young to not realize he was talking to a bot and did something like drink bleach because it told him to.

This was one of us.

He fit lots of behaviors I see here from me and my fellow Lemmy posters. He:

• built computers for himself and family members
• was a hobbyist (at least) coder
• wasn’t a young kid that didn’t know the world. He was 48 or 49.
• was an early adopter embracing the modern LLM technology in 2022 when it first really became public.
• sold his house in an urban metropolis (Portland) and moved to a rural area so he could use his additional wordworking skills on building sustainable housing.
• worked part time at a homeless shelter

Doesn’t this guy sound like someone that would be a Lemmy poster to you too?

He started using LLMs (ChatGPT specifically) as a tool only to advance his hobby and work. When he first started it appears he understood it was just a tool, and didn’t think it was something sentient. Only later after hundreds of hours of exposure did this idea arise in him.

Was there some underlying psychological problem that the LLM exacerbated? Possibly. But at what level was his original underlying issue? Do we all have some low level condition that would make us equally susceptible? I know we’d like to think we don’t, but how do we know? This man certainly didn’t think he did, I’m sure.

Next I think about what it would take for me to get down this bad path without realizing it. At one point would I be talking to a chat bot, not realize it, and let what that chat bot said change or influence my thoughts when I’d have zero knowledge of it being just a fancy program? I consider myself moderately smart with good critical thinking skills, but I’m sure this man did too.

Then it occurred to me that I have to concede that I have, at some point, already interacted with a bot in years past on Reddit or even today on Lemmy and I had no idea it was a bot. Was that interaction a throwaway conversation about pop culture that would have no impact on my world view or was it a much deeper and important political or philosophical conversation that the bot introduced an idea or hallucinated evidence to support a point and I didn’t catch it to challenge it? Am I already a few or many steps down the bad path of falling for illusions of a bot? I certainly don’t think so, but neither did he.

How many of us are already on the same path as this guy and just as ignorant about the danger as the man in the article?

If you play with the parameters you can make all kinds of things happen, but all of those things are still driven by the existing information it already has or can find. It can mash things together in random new ways, but it will always work with components that already exist.

Or purely randomness, but the spirit of your point is sound. And if it is randomness it may be unique output, but the utility of that result may be zero.

There is no awareness of context or meaning that would allow it to make intelligent choices about what it mashes together. That will always be driven by the patterns it already knows, positively or negatively.

100% AGREE. LLMs are not “thinking”. LLMs are NOT the HAL 9000 from the movie 2001: A space odyssey

It’s like doing chemistry by picking random bottles from the shelf and dumping them into a beaker to see what happens. You could make an amazing discovery that way, but the chances of it happening are very, very low. And even if it does happen, there’s an excellent chance that you won’t recognize it.

100% AGREE.

I’m in favor of using LLMs for tasks that involve large-scale data analysis. They can be quite helpful, as long as the user understands their limitations and performs due diligence to validate the results.

Unfortunately what we are mostly seeing are cases where LLMs are used to generate boilerplate text or code that is assembled from a vast collection of material that someone who actually knew what they were doing had previously created. That kind of reuse is not inherently bad, but it should not be confused with what competent writers or coders do. And if LLMs really do take over a lot of routine daily tasks from people, the pool of approaches to those tasks will stagnate, and eventually degenerate, as LLMs become the primary sources of each others’ solutions.

100% agree. The degeneration is already occurring because bad LLM output is being fed back in as authoritative training data resulting in confidently wrong answers being presented as truth. Critical thinking seems to have become an endangered species in the last 20 years and I’m really worried that people are trusting LLM chatbots completely and never challenging the things they output but instead accepting them as fact (and acting on those wrong things!).

LLMs may very well change the world, but not it in the ways most people expect. Companies that have invested heavily in them are pushing them as the solutions to the wrong problems.

I think we have some of the pieces today that will make AI in general more trustworthy in the future. Grounding can go part way to making today’s LLMs more trustworthy. If an LLM claims something as fact, it should be able to produce the citation that supports it (outside of LLM output). That source can then be evaluated critically. Today’s grounding doesn’t go far enough though. An LLM today will say “I got that from HERE” and simply give a document. It won’t show the page or line of text and supporting arguments that would justify its arrival at its stated output. It can’t do these things today because I just described reasoning which is something an LLM is NOT capable of. So we wait for true AGI instead.

LLMs are not capable of creating anything, including code. They are enormous word-matching search engines that try to find and piece together the closest existing examples of what is being requested. If what you’re looking for is reasonably common, that may be useful.

Just for common understanding, you’re making blanket statements about LLMs as though those statements apply to all LLMs. You’re not wrong if you’re generally speaking of the LLM models deployed for retail consumption like, as an example, ChatGPT. None of what I’m saying here is a defense about how these giant companies are using LLMs today. I’m just posting from a Data Science point of view on the technology itself.

However, if you’re talking about the LLM technology, as in a Data Science view, your statements may not apply. The common hyperparameters for LLMs are to choose the most likely matches for the next token (like the ChatGPT example), but there’s nothing about the technology that requires that. In fact, you can set a model to specifically exclude the top result, or even choose the least likely result. What comes out when you set these hyperparameters is truly strange and looks like absolute garbage, but it is unique. The result is something that likely hasn’t existed before. I’m not saying this is a useful exercise. Its the most extreme version to illustrate the point. There’s also the “temperature” hyperparamter which introduces straight up randomness. If you crank this up, the model will start making selections with very wide weights resulting in pretty wild (and potentially useless) results.

What many Data Scientists trying to make LLMs generate something truly new and unique is to balance these settings so that new useful combinations come out without it being absolute useless garbage.

while at the same time, ignoring Windows telemetry,

You’re posting this statement on Lemmy? There is a dispropotionatly high population of Linux and OSX users here. Most of those here ignoring Windows telemetry aren’t running Windows.

He said it was not yet clear how many gunmen were involved, adding that detectives and officers from the Taxi Violence Investigations Unit were investigating the attack.

Taxi Violence Investigations Unit

It also has a good use of being the toilet of browsers. As in, if you ever are required to temporarily install some pervasive plugin or extension to take a proctored exam or something, Edge is good to use because you know you won’t use the that browser for anything you care about and you can protect good browsers from those garbage plugins.

With your comments I found additional German legal guidance that mostly matches what you said. It appears that Germany does indeed have a portion of privacy from someone intentionally walking up to you and taking your picture. I don’t think this invalidates my original point because it doesn’t appear that expectation of privacy extends to installed surveillance cameras in public.

However, I appreciate having a better understanding of German law. Thank you.

Forgive the machine translation to English, but reading that shows the a very similar exception to privacy protection we have here in the USA

Here’s one example:

"There are exceptions to events (demonstrations, general meetings, cultural events, etc.). Here, participants must expect to be photographed. This is about what is happening and not about the person itself. "

Most of the wiki article is talking specifically about copyright, which isn’t the scope of what we’re talking about. Publication of taken images is a different topic.

In my opinion, go the Mondragón route. Bring democracy into the enterprise and allow those who work to control how they work. That way those who are being “automated” away can have a voice in what to do next.

Isn’t that what we already have today? Jim no longer has a job at this employer. Jim can choose where he works next.

Also, your vision of human capacity is very limiting. Why can’t Jim learn new skills? Everyone does it, literally all the time. Even construction workers have domain knowledge on how to pour cement that they learnt from others.

As shown in the example, Jim is not capable of learning the skills (in any reasonable amount of time) to take on another open position at that company. So are you suggesting that Jim go back to school? Who are you suggesting, in your vision, is pay for Jim’s living and school expenses until he is ready to work a position with a higher skillset?

Apathy? Not at all. Its simply a matter of established law, in the USA anyway. I can’t speak to the legal systems of the other 140+ countries on planet Earth.

Can you cite a law in the USA or in your own country where you have a right to privacy making photographing you simply standing in a public park an illegal act perpetrated by another person or government entity?

I think we’re aligned on the core issue but with nuanced perspectives. Regulatory capture is indeed the established academic term for the phenomenon you describe,

Its close, but I don’t think that’s correct for this situation.

precisely capturing how agencies meant to protect public interest end up advancing industry priorities through mechanisms like the revolving doorbetween Boeing and Congress.

You’re missing one key aspect of the definition of regulatory capture. NASA isn’t a regulatory body in the case with Boeing, its the customer.

For it to be regulator capture NASA would have to be acting as a regulatory body, and the corrupt company would have to have influence over policy that they benefit from outside of the regulator. An example of regulatory capture was what lead up to one aspect of the 2008 Financial Crisis. Banks have to have a US government regulatory that sets policy and policies the actions of the bank. Prior to 2008 banks could choose their regulator which their choices between the FDIC, Federal Reserve, or a little known regulator call Office of Thrift Supervision (OTS). It won’t surprise you to find out that the OTS was a tiny little shop which only had a few employees, and banks figured out they could write their own policy, get the OTS to approve it, and get away with actions the banks would normally be barred from doing. This lead to risky bank behavior, and the failure of banks and a large contributor to the Financial Crisis of 2008.

NASA wasn’t acting as a regulator to Boeing for Starliner. NASA wasn’t setting government regulations which Boeing had to follow for all vehicles Boeing produced for spaceflight. NASA was a customer giving specs to its contractor, but the contractor had corporate power over its customer, NASA. So yes this would be something like corporate capture but it wasn’t regulatory capture.

Where I’d argue the Starliner narrative: While Boeing’s participation provided political cover for Commercial Crew legislation,

We agree with this. This was my whole thesis in my original post.

SpaceX’s 2010 Falcon 9 debut and subsequent rapid repeatability fundamentally reset industry expectations.

Not really. It wasn’t SpaceX alone, and it wasn’t because SpaceX as rapid. It was because it was it was cheap. SpaceX wasn’t alone in this though. The other contract winner of Commercial Cargo contract, Orbital Sciences, was also cheap and had nothing to do with rapid repeatability. Both were, however, cheap, compared to the cost-plus contract providers that came before them.

The success of fixed-price cargo contracts demonstrated reusable rockets and rapid iteration were possible, proving cost-plus models weren’t inevitable. This technological inflection point–not Boeing’s involvement–created the political space for NASA to demand accountability in human spaceflight.

I disagree entirely. SpaceX reusabilty had zero impact on the success of the initial Commercial Cargo or Commercial Crew contract adoption. How do we know this? Four ways:

1. When SpaceX started flying cargo, reusuabilty wasn’t even a thing yet on Falcon 9. Reusability arrived later during the contract, but the fixed price contracts had already been signed and SpaceX received no extra money from the contract derived from reusability.

2. SpaceX wasn’t the only provider of Commercial Cargo. The other was Orbital Sciences (later OrbitalATK, later yet Northrop Grumman) with their completely disposable rocket and cargo module (Cygnus). Again, when Orbital signed their contract for Commercial Cargo the prices were set. Whether Orbital threw away their Antares rocket after launch (which they did) or not, had no bearing on the Commercial Cargo contracts.

3. No part of Starliner was reusable at the time of contact signing for Commercial Crew. Not the core stage, not the second stage, not the SRBs, not the crew vehicle. If reusabilty was so much of a factor for Commercial Crew how did Boeing, that had zero usability, not only win a Commercial Crew contract, but also was the highest paid of the two contact winners?

4. If reusabilty was such an important factor in Commercial Crew selection, why was Boeing, with zero reusabilty, chosen, but not Sierra Nevada Corporation’s (today known as Sierra Space) Dreamchaser vehicle NOT chose when it was a reusable crew vehicle from day 1?

Boeing’s Starliner struggles directly stem from its post-1997 merger culture shift,

We agree on all the reasons Boeing sucks today.

The breakthrough came not from Boeing’s inclusion but from SpaceX proving fixed-price development could work

That simply isn’t true. Again, SpaceX wasn’t the only Fixed Price space contractor. Orbital Sciences was too. Also, I remember pieces quotes from government hearings where SpaceX was criticized as not being up-to-the-task of handling human flight and that only a company with experience like Boeing would be able to deliver, and without a “sure thing” delivery contractor extending the concept of Fixed Price contracts from Commercial Cargo to Commercial Crew shouldn’t move forward unless a trusted company like Boeing was involved in Commercial Crew. This was also why Boeing was paid so much more than SpaceX for far fewer flights in the contract language.

Now if they can just notify you that some asshole is recording you on their cell phone instead of reading reddit.

If you’re out in public, always assume you’re on someone’s camera. That isn’t really new either.

before that it wasn’t always considered as big of a deal as you are referring to, idk pre 1970s or what.

We’re agreeing with the reality that it wasn’t considered a crime or a big deal in generations past. Where we have a huge gulf of disagreement is if this was a problem or not. I am flabbergasted about the strong defense you’re putting up to be able to drink and drive.

May I ask if you or your family have ever been negatively affected by a drunk driver before?

I digress though, no one thinks people should be driving drunk, I am just making the point, that .12 for generations was the standard, in some states.

And the standard before .12 was “no standard” where driving drunk wasn’t even a crime.

The larger problem is why we are completely reliant on vehicles, that we cannot even enjoy more than two drinks on the town and legally go home. There must be better ways, fuck cars.

Taxi cabs have exist since before the invention of cars. They were horse drawn carriages. Today we even have Uber and Lyft that are easier that hailing a cab.

Completely unrelated to the article: I would encourage any woman of child bearing age to obtain a passport now when there is no rush. Using the slow process it takes about 6-10 weeks of waiting to get your passport after you apply. For a full passport that can be used in any country the cost is $130. If you only want to go to Canada and/or Mexico, you only need a passport card, which can be had for only $30. Its the same form to get either the book or the card, you would just check a different box.

Also unrelated: Abortion pills are easily available in both Mexico and Canada.

Growing up, our household had a giant roll of butcher paper. It was 2 ft (60cm) wide and about 1000 feet (300m) long roll. I have no idea why we had it, but as kids we were allowed to use as much as we wanted for whatever we wanted. It turned into a childhood of projects, games, costumes, banners, signs, crafts, wrappings, pranks, etc. Close to the beginning as kids, we’d asked for art supplies like markers, paint, pens, pencils, charcoal, etc to transform that boring cheap paper into different universes. We became creative because it was available.

Something about having an unlimited supply of something and infinite permissions was an unexpected freedom.

Uh huh, hey, why don’t these job numbers reports ever talk about whether these new jobs are keeping up with the cost of living? Seems like it’d be important to discern how many jobs are paying minimum wage and how many are paying enough to actually afford to survive longer than the next 24 fucking hours.

You’d get closer to that answer with a different report. Probably a combination of the Occupation Finder data showing wage ranges and the Employment Projections data which shows employment increase in number of jobs or declines in each sector.

The BLS used to be a gold standard for fantastic data collection, analysis, and sharing. However, I am not putting much confidence behind any data coming out of the trump administration.

I have no problem with Bad Bunny’s music or entertainment even if its not something I seek out regularly. However, I’ve never seen Bad Bunny work against free and fair elections in the United States or foment insurrection when he didn’t get elected President, so just by those measures, Bad Bunny is a much better representation of American values than trump.

Is it surprising though? Those that put more thought into it choose not do to the violence.

Its not an equally slow CPU. These boards support Xeon CPUs that first launched 11 years after DDR3.

The implication is that there are users needing large RAM footprints that aren’t CPU bound. The hit in performance on the RAM isn’t significant enough to justify spending orders of magnitude more for modern DDR5 which is in short supply.

In computing history there’s precedent for this. Amiga computers had a small amount of “Fast RAM” which was extremely expensive, but the CPU could address a second bank of “Chip RAM” which was significantly slower but much much cheaper.

We could see this idea return in modern computers.