nhamiel.bsky.social - Profile | ThreadSky | a Reddit-style client for Bluesky

My new piece in @theguardian.com Techno-optimism is human pessimism. www.theguardian.com/commentisfre...

submitted 4 days ago • 16 comments

#FountainPenFriday A Pilot Silvern Sterling Silver Dragon with a broad nib. Not for people easily annoyed by smudges or fingerprints :) A lovely, smooth writer.

submitted 3 days ago • 0 comments

Who can spot the issue with this paper? Don't worry, it's visible from the introduction. genaisecurityproject.com/resource/owa...

submitted 3 days ago • 0 comments

Don't allow AI to wallpaper over BS. This isn't what's happening here. People making these comments don't know what they are talking about and have never even touched these tools, much less know their current capabilities and limitations. www.csoonline.com/article/3829...

submitted 4 days ago • 0 comments

While being so distracted by AI, people have totally missed the fact that mullets are making a comeback.

submitted 4 days ago • 0 comments

“Can you look at this RFP? What do you think?” Hmmm… It looks like they are attempting to select the best wand in a fantasy world where magic actually exists, threatened by a famished demon salivating over their souls. “Got it, thanks.” 😆

submitted 5 days ago • 0 comments

Call me old, but I remember when apps kinda… I don’t know, worked? They had a certain level of reliability and consistency, at least. Modern apps have become a bit of a dumpster fire. No wonder we’re willing to accept wonky AI outputs as acceptable.

submitted 6 days ago • 0 comments

In the end, human harm results not from intelligence but incompetence.

submitted 6 days ago • 0 comments

It’s 2025, maybe don’t continue to fall for literally EVERY demo. Haven’t people learned anything from the past couple of years? Fool me once, shame on you, for me 10,000 times…

submitted 10 days ago • 0 comments

The Slop Generation won’t be defined by cultural constructions and artistic expressions but by the output of one-arm bandit slop machines where artistic expressions aren’t created to be admired but as individualized wallpaper blending into the background noise of life. perilous.tech/2025/02/20/t...

submitted 11 days ago • 0 comments

This from the CEO who said he thought AI would replace him too. Quite a few people sent me the Klarna example as though it were proof I was wrong about something, but this is inevitable when so little of the problem space is considered, along with the devaluation of what people actually do.

submitted 11 days ago • 0 comments

Apparently, I’m a victim of some geoengineering conspiracy theory named Fogvid-24. I haven’t noticed the chemical smell, but I have been sick for a few weeks now, so I’m not sure if I should blame aliens or humans 😆 futurism.com/florida-myst...

submitted 17 days ago • 0 comments

Very cool work from the team at Meta. A more comprehensive and valuable approach than purely asking an LLM to write a test. engineering.fb.com/2025/02/05/s...

submitted 18 days ago • 0 comments

Quite a few people working in tech think people who don’t shouldn’t have jobs. These same people think programs that don’t lead to technological progress are a waste of money. If you are trying to understand the current moment, look no further.

submitted 18 days ago • 0 comments

I am a product of Head Start. I certainly understand the value firsthand. However, I fear that these programs will be written off as pointless wastes of money because people who don’t come from a low-income background don’t see the point. apnews.com/article/head...

submitted 19 days ago • 0 comments

So many people believe that AI will solve their problems, not realizing that these very problems need to be solved before they can successfully apply AI. It’s like trying to throw a large dinner party at a hoarder’s house. It’s not going to work.

submitted 19 days ago • 1 comment

Are you interested in submitting to the Black Hat AI track? Have a look at these observations and recommendations to ensure your submission covers the bases. Don’t hesitate to reach out if you have any questions. perilous.tech/2023/10/04/b...

submitted 26 days ago • 0 comments

There’s a fallacy that in the future, things will cost nothing, and everyone will live a life of luxury if only we let companies raw dog technology all the way to utopia. This is a smokescreen for unknowns and outcomes that won’t lead to luxury. perilous.tech/2025/01/31/t...

submitted 31 days ago • 0 comments

LLMs are handy for a bunch of tasks, but if you use them to replace your thinking rather than improve it, you're only hurting yourself in the long run. Good long-form writing is hard. It's mental exercise. Having an LLM half-ass it for you robs your readers and yourself. Don't skip leg day.

submitted 31 days ago • 1 comment

WTF is wrong with these people? By the way, this is NOT how you debunk child blood conspiracy theories.

submitted 32 days ago • 0 comments

I must have missed the point where immortality was added to my Claude subscription. You know LLM hype is out of control when people claim they can make us immortal.

submitted 34 days ago • 0 comments

I guess this was left in for ironic emphasis? Having a hallucination in a document that also warned about the risks of hallucination.

submitted 34 days ago • 0 comments

The market reaction to DeepSeek is off if the goal is to create AGI and not just shove LLMs into everything. LLMs are not a path to AGI, so if you are investing in the development of AGI, the data centers and other investments are probably still necessary.

submitted 34 days ago • 0 comments

After users started reporting receiving 60-minute unskippable ads on YouTube, Google sent a statement to journalists that the company is dead serious about going about ad blockers 😂 www.androidauthority.com/youtube-long...

submitted 35 days ago • 3 comments

The Four Ds of Personal #AI #Risk. Although we don't have true personal AI systems today, they're on the horizon—much sooner than AGI. Understanding the risks is the first step to protecting what you care about. perilous.tech/2025/01/27/f...

submitted 35 days ago • 0 comments

The gauntlet of inefficiency imparts discoveries that optimization destroys. Some things aren't meant to be easy, and the value of the activity is generated from the friction incurred. This is especially true in things like writing and art.

submitted 38 days ago • 0 comments

500 billion

submitted 40 days ago • 0 comments

Wait, what??? 🤣

submitted 42 days ago • 0 comments

here's the post vickiboykis.com/2025/01/14/h...

submitted 48 days ago • 3 comments

No... Get outta here. Say it ain't so 😆

submitted 46 days ago • 0 comments

I collected some thoughts on being realistic about agents in 2025. It’s important to consider that agents can resemble grenades when deployed in high-risk settings, where the damage doesn’t happen the moment you pull the pin. 2025, here we go. perilous.tech/2025/01/13/b...

submitted 49 days ago • 0 comments

He's going to be known as the little boy who cried AGI.

submitted 52 days ago • 0 comments

The article I wrote for the WEF on deepfakes was published today. For all the hype, hand-waving, and fire-setting about the danger of deepfakes on elections, they were a dud, and yet so many were so sure they’d create a misinformation apocalypse. www.weforum.org/stories/2025...

submitted 52 days ago • 1 comment

Quite a few tech people think most people don’t seem to do any work, devaluing jobs they don’t do or don’t directly relate to production. A special forces unit can’t fight without the food service staff. “Most cooks are a form of UBI. They come in, make three meals, and leave.”

submitted 54 days ago • 0 comments

Radio! With its highly customized content, algorithmic curation, and infinite scroll. The amount of brain rot that manifested from the radio was mind-boggling. Radio was a one-way device that was impossible to cognitively offload to. So, yes, exactly the same. 🙄

submitted 55 days ago • 0 comments

Hype. It’s that simple. It’s 2025. Let’s make it the year to add at least “some” skepticism, not believing every claim or demo as though it’s gospel.

submitted 56 days ago • 0 comments

I'm going to finish this song before my vacation is over :)

submitted 59 days ago • 0 comments

You are going to hear the word “agent” so many times in 2025 that you’ll never watch another spy movie again.

submitted 59 days ago • 0 comments

Nils’ talk, AI Meets Git, presented at CCC is now available to stream. This talk covers still unpatched vulnerabilities in Qodo Merge 🙄 Let’s say it again: changing the name of your company and product isn’t a valid patching strategy. media.ccc.de/v/38c3-ai-me...

submitted 63 days ago • 0 comments

The shift from assistant to companion will be driven by a push for product stickiness and access to more data, with the potential for manipulation in the future. perilous.tech/2024/12/30/t...

submitted 63 days ago • 0 comments

I checked in on Friend.com today. It’s going as well as expected 🤣 No, I don’t want to take Craig anywhere! All he does is act dramatic. Just what humanity needs: venting, whiny bots complaining about events that never happened.

submitted 66 days ago • 0 comments