@Gaywallet@beehaw.org avatar

Gaywallet

@Gaywallet@beehaw.org

I’m gay

This profile is from a federated server and may be incomplete. Browse more on the original instance.

Gaywallet ,
@Gaywallet@beehaw.org avatar

Fantastic article highlighting the issue. Thanks!

Gaywallet ,
@Gaywallet@beehaw.org avatar

Stop arguing about what slurs are okay to use. The only rule around here is to be nice. If someone asks you to not use a word because it hurts them, the nice thing to do is to listen.

Gaywallet OP ,
@Gaywallet@beehaw.org avatar

Wow, what a constructive and useful comment. Thank you for contributing 💜💜

Gaywallet OP Mod ,
@Gaywallet@beehaw.org avatar

Simplifying this down to an issue of just the review process flattens out the problem that generative AI does not think in the same way that generative human content does. There's additional considerations that need to be made when considering using generative AI, namely that generative AI does not have a sum of knowledge to pull from in order to keep certain ideas in check, such as how large an object should appear and it doesn't have the ability to fact check relevancy with other objects within the image.

We need to think about these issues in depth because we are introducing a non-human, specific kind of bias into literature. If we don't think about it systematically we can't create a process which intends to limit or reduce the amount of bias introduced by allowing this kind of content. Yes, the review process can and should already catch a lot of this, but I'm not convinced that waving our hands and saying that review is enough is adequate to fully address the biases we may be introducing.

I think there's a much higher chance of introducing bias or false information in highly specialized fields where the knowledge necessary to determine if something was generated incorrectly, since generative AI does not draw upon facts or fact check, is in fact, correct. Reviewers are not perfect, and may miss things. If we then draw upon this knowledge in the future to direct additional studies we might create a house of cards which becomes very difficult to undo. We already have countless examples of this in science where a study with falsified data or poor methodology breeds a whole field of research which struggles to validate the original studies and eventually needs to be retracted. We could potentially have situations in which the study is validated but an image influences how we even think (or can acquire funding for) a process should work. Having strong protections such as requiring that AI images be clearly notated that they were created via AI, can help to mitigate these kinds of issues.

Gaywallet OP Mod ,
@Gaywallet@beehaw.org avatar

I am most certainly not waving hands and saying that review is enough

Apologies, that's what it sounded like to me. You said it's clickbait. You said the title would work without AI in the title. You also said that AI generation isn't relevant. That felt like diminishing the conversation - focusing in on what you're most concerned about, and dismissing all other discussions. I don't think that helps discussion happen. It discourages it. It says that we shouldn't talk about the problems present here which exist outside the realm of just the review process.

For example, both of the figures do have a description, but neither of them have any kind of attribution. The review process might ensure it is factual when it is followed and still let through material such as that you've laid out above which do not involve AI - like hiring someone off of fiverr. One way to solve this would be with image attribution. As I mentioned above, simply requiring that an image explain where it came from, such as requiring attribution to the artist who created the figure or requiring that the software used be attributed (perhaps even requiring the full prompt for generated images) are all methods through which we can ensure scientific rigor (and accurate attribution) which will both help ensure the review process catches problematic material and cues the readers in to key information about the figures present in research.

Gaywallet OP Mod ,
@Gaywallet@beehaw.org avatar

https://beehaw.org/pictrs/image/a732ddfa-0a60-4d82-b5d3-93da88ae7d1d.webp

Direct link to journal article -https://www.frontiersin.org/articles/10.3389/fcell.2023.1339390/full

Gaywallet ,
@Gaywallet@beehaw.org avatar

Insulting people isn't nice. The only rule on this understand is to be(e) nice. This is not appropriate behavior for our instance. You didn't need to insult someone to make your point. I'm giving you a 7 day ban to think things over.

Gaywallet ,
@Gaywallet@beehaw.org avatar

Very approachable article for people who don't have experience with the fediverse. Thanks!

Gaywallet OP ,
@Gaywallet@beehaw.org avatar

It appears to have been lazy reporting. Didn't happen

Gaywallet OP Mod ,
@Gaywallet@beehaw.org avatar
Gaywallet OP Mod ,
@Gaywallet@beehaw.org avatar

I can relate to being vulnerable and extremely influenced by factors outside of my control, tho

Gaywallet OP ,
@Gaywallet@beehaw.org avatar

I think the most interesting finding in this study is the following:

The models also suffered from “contra-factual bias": They were likely to believe a false premise embedded in a user’s question, acting in a “sycophantic” way to reinforce the user’s mistake.

Which when you think about how language models work, makes a lot of sense. It's drawing upon trained data sets that match the question being asked. It's easy to lead it to respond a certain way, because people who talk pro/con certain issues will often use specific kinds of language (such as dog whistles in political issues).

Microsoft stole my Chrome tabs, and it wants yours, too (www.theverge.com)

Last week, I turned on my PC, installed a Windows update, and rebooted to find Microsoft Edge automatically open with the Chrome tabs I was working on before the update. I don’t use Microsoft Edge regularly, and I have Google Chrome set as my default browser. Bleary-eyed at 9AM, it took me a moment to realize that Microsoft...

Gaywallet ,
@Gaywallet@beehaw.org avatar

One thing though: Criticism of admins should never be considered a rule breaking event provided it is not derogative or endangering, and if my reply to you is considered a reason for admin action then I need to reconsider my participation in beehaw as well.

Wanting to block an admin isn't criticism. You're free to criticize us, and in fact, encouraged to do so if we are warranting criticism. If you were able to block us, we wouldn't be able to tell you when you are breaking the rules or provide feedback based on reports of your comments. We can't run a space like this without giving this kind of feedback to members, and just like what happened in this case, we try to give this feedback before jumping to moderator actions like banning.

Gaywallet ,
@Gaywallet@beehaw.org avatar

Let me be clear: The ONLY site I have ever been banned from was reddit for posting ‘Punching nazis is a moral good’, and please tell me now if this is a violation here and I will just delete my account and leave.

Punching nazis is most definitely a moral good. We've outlined this (intolerance towards the intolerant) in our docs.

You're Not Imagining It: Google Search Results Are Getting Worse, Study Finds (gizmodo.com)

For the past few years, a growing number of users, analysts, and experts raised alarms about a truth that feels obvious to a lot of people who surf around in web browsers: the quality of Google results is in serious decline. Google disagrees.

Gaywallet ,
@Gaywallet@beehaw.org avatar

As an FYI, opposed does not necessarily mean opposite, it can and often means in contrast to or in conflict with. Shades of gray, but either word works fine here.

Gaywallet ,
@Gaywallet@beehaw.org avatar

Yikes, thanks so much for these links. I've avoided kagi for awhile now for lack of need but it had been top of list of 'things I should try'. Guess I can strike that one out now.

Gaywallet OP ,
@Gaywallet@beehaw.org avatar

You're absolutely correct, yet ask someone who's very pro AI and they might dismiss such claims as "needing better prompts". Also many people may not be as tech informed as you are, and bringing light to algorithmic bias can help them understand and navigate the world we now live in. Dismissing the article just because you already know the answer doesn't really encourage people to participate in a discussion.

Gaywallet OP ,
@Gaywallet@beehaw.org avatar

This is incredibly dismissive of the concerns raised and adds nothing to the discussion

Gaywallet Mod ,
@Gaywallet@beehaw.org avatar

I get what he's saying, and in many ways I agree, but the choice of words is too strong for the hypotheses he's presenting. For example, he uses the following to bolster his claim

Abundant evidence indicates that people who grew up in homes marked by chaos and deprivation will perceive the world differently and make different choices than people raised in safe, stable, resource-rich environments.

Yet he mentions himself that we are subject to our external environment. Some of these individuals do not make markedly different choices based on these external differences, not to mention their own internal ones (genetics, etc.).

To make the claim that we have no free will because we are the sum of our environment + upbringing ignores that we have a modicum of control over our environment, and it also ignores how our interactions provide that external environment. We pass laws to further human rights and create a better environment so that people in the future would hold them in higher esteem and be influenced by these choices we make. In short, there is a field of possibility that lies within the maybes - our genetics and upbringing set us up for how malleable we are on any decision. Some decisions simply won't happen and some outcomes are likely inevitable, but most fall in the space where there is a likely but not predetermined outcome which is influenced by the environment. This is why perfect predetermination is impossible.

In fact, this very viewpoint is even reinforced in the most physical of sciences - physics. In quantum physics we can at best determine probable outcomes. While there are theorists who believe in superdeterminism, or the idea that we simply don't have all the variables to determine everything perfectly yet, superdeterminism has gotten no closer to explaining bell tests, the slit experiment, or other quantum phenomenon in well over 50 years. Increasingly complicated mathematics repeatedly show that there is a fundamental randomness to the universe that we seem unable to capture.

And I think it makes sense, in the context of what we know of biology and evolution. Brains are constructed in a way where signals are created almost randomly, and then organized to make sense of the world. Evolution has played a part to refine this processing so that it ignores what's not important to survival and proliferation. If this process weren't generative and random in some sense, we would not evolve and there would be little to no purpose for diversity. The world is constantly changing and thus our biology must account for this, meaning that it must be malleable and open to changes by the environment. If it is open to changes by the environment, then we must be able to influence each other and thus a concept of free will must exist that at the very least is a representation of the sum of all that activity.

Gaywallet OP Mod ,
@Gaywallet@beehaw.org avatar

I spoke with the researchers on this particular study a few months ago at a local conference. This study is interesting to me due to a few factors:

  • Ibogaine is almost never studied in the medical context because a small number of individuals experience an adverse reaction involving the heart which can result in death
  • The researchers theorized you could counter this rare but life threatening side effect by co-administering magnesium, suggesting that afib is behind the heart-related problems and that ibogaine may cause some kind of imbalance with Ca, K, and Na.
  • Ibogaine is a schedule 1 controlled drug in the US meaning that it cannot be studied for medical purposes. Unlike MDMA and psilocybin which have had the power of MAPS arguing for decades for the use in research and allowing medical research despite the scheduling status of the drugs, ibogaine has not received this special status, meaning that the researchers had to have a rather unique study design in which patients were recruited in one country and sent to another country for treatment. I've quite literally never seen a study do this, let alone one which is working with a population that is federal in nature (veterans) and is a fantastically creative way of ensuring medical research can continue amongst draconian law.

Also, here's a direct link to the article in nature

Gaywallet ,
@Gaywallet@beehaw.org avatar

I used to think that SEO only affected people trying to sell products directly. It's sad how badly it's polluted the ability to find any content at all on the internet, as it's cheap enough and easy enough to do at this point that they're just trying to sell on ad revenue/clicks.

Gaywallet Mod ,
@Gaywallet@beehaw.org avatar

Both happen! It's just work on phage therapy had to reinvent itself in the last decade or so and modern techniques are only just reaching maturity. We're gonna see a lot more development in this space over the next decade.

Gaywallet ,
@Gaywallet@beehaw.org avatar

David Bronner, one of the grandsons of Dr. Bronner, is a fantastically interesting person with a lot of really charming eccentricities I think you were trying to capture when you said 'crazy talk'. I've met him a few times at medical psychedelics conferences, he was introduced to psychedelics at burning man and has bought fully into how it's expanded their mind and has helped them heal.

Gaywallet ,
@Gaywallet@beehaw.org avatar

I would say that's their story to tell

  • All
  • Subscribed
  • Moderated
  • Favorites
  • random
  • All magazines