Tag: ai

Some Thoughts On AI, The News, and Humanity

Carl Setzer4 Comments
bionic hand and human hand finger pointing

Meta has had two significant issues brought up this week (hat tip to Futurism):

  1. Facebook Is Hosting Huge Numbers of Horrifying AI-Generated Videos of Violent Child Abuse, and Meta Is Barely Even Pretending to Care About Taking Them Down
  2. Mother Horrified When Meta’s AI Scours Instagram and Facebook for Invasive Information About Her Young Daughters

Some quick observations:

re: #1: Meta’s PR team sent this note to the Futurism team: “Not all the links you sent violate our policies,” it read. “Can you please avoid suggesting in your reporting that they do?” My main thought isn’t “why doesn’t this violate your policies?” This is horrifying content. But it is not CSAM. Which is what has the attention of government, the public, etc. I’m not too surprised that Meta was caught off guard about both the existence of this content and its prevalence. I also expect that their “Community Standards” are much more reactive. Therefore, since they didn’t anticipate content featuring physical violence against children, they haven’t created a community standard to be violated. I expect that will be fixed “soon”, when the committee assigned to evaluate this issue finally finishes their review. This game of whack-a-mole with humanity’s darkest impulses is troubling, and I wonder if there’s a way to be proactive and not dive into Minority Report territory.

re: #2: This story is troubling in a different way. AI’s ability to analyze a monumental array of data and ACCURATELY detail someone’s personal details is amazing from a technical perspective, and terrifying from a privacy one. As I was reading this article, I came up with the phrase “stalker-tech” to describe it. Now, before we come swinging at Ms. Robins for sharing ANYTHING about her children, she’s not some influencer sharing her intimate details. Most of the information harvested was shared more privately, amongst friends and family. Meta’s AI triangulated amongst a huge spread of information in order to derive deep details about these children. On an aside: I think her response of asking folks to take down these images is, well, hopeless. The information has been harvested. Removing elements of it from one place on the internet doesn’t delete the graph. Meta’s response? “But we missed the mark because the feature never should have prompted the individual with questions like that and we’ve fixed that issue.” I’m curious about how the “issue” has been fixed. Will the prompts not ask similar questions? Is the ability to analyze huge swaths of data been minimized? My guess is that the prompting has been modified, but that the data is still mapped. Therefore, I expect this sort of misuse is still possible, it will just take more steps and bad intent to execute on it.

Humans are a reactive lot. We’re terrible at thinking technology’s negative possibilities through before we run into problems/crisises. We tend towards letting the potential positives blind us to any negatives. I wonder if we have any hope of changing that part of our nature.

    Now We’re At Risk Of AI Swarm Attacks

    Carl Setzer1 Comment
    man and woman with prosthesis in futuristic surroundings

    I just read ‘Sophisticated’ AI swarm attacks are months away, OpenAI warns: What experts say businesses must do and now have a solid guess of what I’ll be coping with the rest of 2026. Whee!

    So, as we’re cataloging all the recent rogue AI attacks, things are becoming sobering.

    • Open AI’s attack on Hugging Face is a big one
    • Anthropic had one “escape”
    • Meta
    • Grok

    The list is amazing and a bit frightening. Reading the article will give you a solid intro to the key attacks to date.

    There are some things those of us in IT need to be considering in order to harden our network’s defenses, too. The article gives some basic insights.

    But, really, what are WE, as individuals, to do? The article points to a couple of things:

    • Move to passkeys for your login credentials.
    • Do NOT reuse passwords!
    • Use a password manager

    These things are good starting points. Be watching for more news over the coming days. Do watch the ones that are tinted with hysteria, though.

    And yet more thinking about AI

    Carl Setzer1 Comment
    person reaching out to a robot

    This morning, I read a LinkedIn post by Bryan Creely, who writes about career management and job hunting. (I highly recommend his work, btw).

    I feel that this quote summarizes the whole thing:

    They spent the last 3 years scaring us into believing we were about to go the way of the dinosaur.

    And now they’re walking back that narrative and beginning to rehire for roles they eliminated due to AI.

    That’s what pundits do, prognosticate, then shift, following the zeitgeist where it leads. That’s the case with AI. When they were selling AI to corporations and investors, it was all about replacing workers. That most jobs were going to be replaced by AI. But, anyway, as I expected, AI’s costs turned out to be more than estimated. I believe the abrupt rise in costs caught many companies completely off-guard, though I don’t think it was hard to predict. I wasn’t exactly right on my reasons, but I expected that token costs would have to go up. So, yay me.

    Not too surprising, but we haven’t heard these pundits acknowledge their error or overestimation. Yet they are quietly walking back a lot of the hype. Better than a bursting bubble, though I’m not sure we’re out of the woods on that possibility yet.

    One additional thing: I think those hyping AI were stunned by the way the public went negative on them. They seemed truly stunned by crowds of college grads booing them, by protesters at hearings for proposed data centers, and by the bipartisan nature of it all. But, really, your messaging leaned on the redundancy of most of the working population, and not-so-subtly hinted at mass unemployment and other such things. Add in electricity costs, and how the price increases are being pushed onto the communities these things are in, environmental costs, and it is so obvious that this was going to be the outcome.

    So, it makes me feel better that we’re stepping back some from the irrational exuberance before this thing exploded. Yeah, I’m not convinced we’re free of the bubble collapsing, though my hope is that it’ll gently fizzle and deflate. I can be hopeful, after all.

    Let me end with this comic from Pascal Bornett.

    I’m fascinated by this AI music video

    Carl Setzer2 Comments
    stylized ai typography on pink background

    I often listen to these AI-generated tracks. I find them soothing. Many of the videos, though, are wild in their AI mistakes/hallucinations. I’ve seen boats and airplanes zooming backwards, people walking through walls, each other, all kinds of things. At least once, a person launched upwards into the air, up into the sky. Now, this one below is interesting as the subject remains pretty still (he is supposed to be a whirling dervish, though), but this cloud of sand billows out from his stationary feet. Besides that, I find the colors rich and vibrant, the details nice, but the billowing sand is…odd.

    Anyway, I hope your Wednesday was fantastic and that your June is off to a good start.

    Are we nearing the end of AI’s hype cycle ?

    Carl Setzer3 Comments

    I just watched this video and it got me thinking about the future of tech. One of the key things they point out is how many of the AI prognosticators are now so walking so much of it back. Companies are hiring back some of those laid off. The drive to push everyone into using AI for everything is hitting up against costs, as tokens are no longer free, or subsidized. Anyway, check it out! It’s only about 10 minutes long.

    I’ve written about AI before, and about how the current costs are not the true costs. That much of the operational costs of AI have been underwritten by investors. And that’s not tenable. I’ve been seeing more and more articles about these costs being passed long to users. I’m seeing companies burn through yearly token budgets in the first quarter of the year. I’m seeing companies pull back on the requirement to use AI. Yeah, the hype seems to be waning.

    So, are you seeing the same signs? Let me know what you’re seeing.

    Claude’s Mythos and AI Magnified Fears

    Carl Setzer6 Comments

    I missed the intro of Claude’s Mythos, and only started hearing about them in the ensuing panic. This article has a good rundown of the issues that have the cybersecurity world nervous. I do recognize that this could simply be hyperbole meant to bring extra attention (marketing) to the product, but I thought it worth looking into deeper.

    Anthropic just made AI scarier

    Update:

    The Vox article above is (inconsistently) paywalled. I was able to get back and view it one more time before it locked me out again. So, let me give a quick summary of the concerns here.

    “Claude Mythos Preview was designed for general use, Anthropic says, but during testing, the company found it extremely effective at identifying vulnerabilities in the security systems of all types of software, creating potentially massive security concerns.” So, Mythos is the ultimate hacker.

    With this in mind, Anthropic has limited Mythos to a handful of tech companies and banks (Project Glasswing), with the hope that those orgs will be able to shore up their defenses before Mythos exposes their vulnerabilities.

    “How is Anthropic choosing who to share this technology with?

    They’re essentially looking for cyber defenders or companies that a lot of people depend on, and that downstream it would be a huge issue if they got hacked in any way, shape, or form.” It looks like the folks at Anthropic are deciding for themselves, which is all well and good. I think it would be good to slowly roll it out to other orgs, for the same reason. Why is Chase allowed to patch their vulnerabilities while my local credit union is not?

    Anyway, these tools are more effective than we think, and in ways we have not imagined. I’m glad that, at least in this instance, we had the ability and wherewithal to be cautious.

    Fun With AI: “Smart Ass”

    Carl Setzer3 Comments
    fun with AI

    I’m continuing to explore AI. I haven’t done much with image creation yet, though I have done some.

    Came up with this idea after making a smart-alec comment to a colleague. Just a natural flow from there. Simplest of prompts: ‘Please create an image of a donkey in a lab coat lecturing college students, label it “smart ass”‘.

    Yes, it’s silly, but shows the power of these tools. Maybe a minute of effort.

    PS, that includes the “featured image” on this post.

    Anyway, AI is a critical tool and not going away anytime soon. I intend to study it and understand both its strengths and weaknesses. It will impact so much tech, from coding to network management. I’m doing my best to be proactive, not reactive, to emerging tech.

    A Critique On AI Hype

    Carl Setzer8 Comments
    hand of a person and a bionic hand

    AI is seemingly everywhere right now, and I’ve written a bit about it before. As I’ve started diving more deeply into it, I see all the hype about how AI will be disrupting work. Well, I’ve developed a somewhat contrarian viewpoint to the prevalent viewpoint of “AI taking over everything”. Here are a few random thoughts on the whole thing right now.

    A lot of what AI is replacing now, and expected to in the near future, runs on the assumption that it will remain (nearly) free. However, most of the current costs are funded by investor dollars. As MIT recently reported, 95% of AI initiatives are failing to generate value. So, the only way companies are gaining value from AI is via its low cost. And the operating costs are significant. Data centers are expensive, both to build and to operate. GPUs aren’t cheap to buy. The environmental costs are steep. And, right now, those costs are being paid from investor dollars. When those dry up (investors need to make a return on that investment at some point, and that needs to be greater than the investment as well as the current operational costs. I expect the point when investors stop feeling FOMO and start worrying about making a return will be coming soon. I expect that when the costs become the responsibility of the consumer, the price will escalate abruptly and steeply. Will a junior developer really be displaced by ChatGPT or CoPilot? Or will we suddenly find junior developers much more cost-effective?

    Adding to this, I’ve seen discussion about how AI infrastructure investment is potentially masking a recession. Thus, I’m concerned that we’re missing the weakening economy, things like continued tech-sector layoffs, and are not paying attention to some big economic weaknesses. If AI investment dries up, and layoffs start compounding, the dominoes start to fall, and that brings a significant recession/depression. Which will muddle this significantly.

    Anyway, I think that we are in a bubble regarding AI. Like the Dot Com bubble, I think that AI will herald significant technological changes. AI is a powerful tool, and will continue to be so. However, I am confident that the techno-dystopia predicted by AI hyper-meisters will not materialize. I’m not sure we can even conceptualize what it will look like at the end of this cycle. But I am confident that those who have mastered AI will benefit greatly from the coming revolution. There are opportunities, and now is the time to make the investment in understanding them.

    An AI Manifesto

    Carl Setzer5 Comments
    bionic hand and human hand finger pointing

    Cassidy Williams is a tech blogger I’ve followed for years. Recently, she wrote about her views on AI Usage, which I find exceptionally relevant right now. She references the “AI Manifesto” by Damola Morenikeji. Damola suggests creating an /AI section on your page, which I think I’ll do soon. But I’m just starting the exploration now.

    I believe the transparency around our respective use of AI is critical. Being able to trust what you see is becoming harder.

    Me? I’ve used AI as something of an editor for this site, and have explored using it to create pages. I’ve never been 100% satisfied with what Gemini or ChatGPT has created, but I will say it’s carved my workflows significantly. I don’t envision ever wholly turning things over to AI, and see the possibility of adopting Cassidy’s attitude of “Everything on this website is written by me, personally, not an AI.” This quote she shared really gives me pause: “If you couldn’t be bothered to write something, I won’t be bothered to read it”. Yeah, I feel that. I know there’s a reason it’s called “AI slop” and the last thing I want to do is contribute to the enshittification of the internet.

    So, have you created a “Slash AI” page? If so, have you shared it to this list (https://slashai.page/)? If not, but you want to, submit it here. Are you planning on it? Let me know! I’m intensely curious.

    AI Impacts On Admin Work

    Carl Setzer1 Comment
    woman in professional wear seated in front of monitor

    I’ve heard a LOT about the expected impacts of AI on developers. When I was laid off right at the beginning of the pandemic, I opted to make a career pivot into Web Development, so I’ve paid attention to this trend closely. Add in all the lay-offs within the whole developer community, and this career, so recently highly lucrative, has become rather bleak. So much so that I’ve been looking backwards at returning to admin work (which is one of my two jobs right now). Now I see that admin work will be getting hit pretty hard by AI, too.

    Earlier this week, I was on a Google Meet. With that, we used the built-in tool to create meeting notes. And, my God, it worked nicely! I was quite impressed. Now, as a long-time admin, taking and distributing meeting notes has been a significant part of my work. I’m confident that this will be done by AI going forward. And I expect that transition to be pretty quick.

    Another key part of being an admin is scheduling travel. I expected AI to be able to coordinate travel pretty well. So, I did an experiment where I gave ChatGPT a very rough itinerary for a multi-stop business trip, asked it to recommend flights and hotels. It did a nice job with a few extra prompts. I had a pretty solid itinerary within 2 minutes. Impressive.

    After just these two considerations, I am confident that AI will revolutionize Admin work significantly. With the job outlook for this work bleak (BLS projects a loss of 12,400 admin jobs between 2024 and 2034), and AI eroding the work, I expect it to become harder to find roles. (Yet it’s the main type of work I’m recruiters reach out to me for…go figure). It makes my decision to pivot to web development during the pandemic, and now shifting to data analysis seem that much more prudent.

    Anyway, onwards!