mapmeld

Thanks for sharing it. Everyone who thinks their individual DNA data is valuable has yet to explain why 23andMe was going bankrupt.

> it is equally naive to say with certainty that it has not “awakened” either

Have to disagree. If someone says their dog "awakened" with superhuman math skills, or that my local LLM felt hours of loneliness while I'm sleeping and not talking to it, we would recognize the first as a delusion, and the second as a technical misunderstanding. Even though I don't fully understand what is going on in a dog brain or LLM.

I agree it is difficult to say what an LLM is doing, but most of these stories jump to saying it has humanlike experiences, or acting like it has an always-on consciousness. So instead of seeing the emergence of a remarkably unfamiliar and not-human consciousness, it is falling into familiar sci-fi tropes. The Geoff Lewis posts were pretty directly following the writing style of the "SCP Foundation" collaborative fiction.

Before a recent annual performance review, I looked over my stats in the company monolith repo and found out I was net-negative on lines of code. That comes mostly from removing auto-generated API code and types (the company had moved to a new API and one of my projects was removing v1) but it was quite funny to think about going to work every day to erase code.

I thought the latest on this was that the journalist's number was in an internal email from spokesman Brian Hughes, and software or human error led to his phone number being associated with Hughes in Waltz's phone contact

When I've seen discussion of these in the past, belief is often connected to fear or guilt stemming from how bodies react differently to exposure and/or infection. There's an element of randomness whether you breathe in enough of a pathogen, or run into a super-spreader. It can be a relief to think that people are predisposed to do better or worse, and either you were already immune or you can control it.

At scale, there are a lot of confused people who do unexpected stuff. The maintainer of cURL has people contact him when a notice shows up in car software or when they think he is connected to hacking: https://daniel.haxx.se/blog/2018/02/16/why-is-your-email-in-... https://daniel.haxx.se/blog/2021/02/19/i-will-slaughter-you/

I have a similar developer-without-college background and did a semester at WGU to see if that would jumpstart things

First, you have to take several general education courses and complete them alongside your technical courses. If you get stuck on an essay or exam you can't skip ahead and finish Python and JavaScript intro courses

Second, CompTIA exams built into the program cover a lot of Windows and IT security group stuff which I never encountered before as a web / software dev. Just like when I was in school, I found it difficult to memorize their definitions of tech terms and take the exam seriously (after all I've been fine without it). The cert exam was monitored by a remote proctor who needed to see my whole room. So I'd recommend going through a certification first and see how that makes you feel.

This detail about the Smithsonian was news to me so I went reading up on it. Curtiss had been involved in supporting Samuel Langley as the key breakthrough in aircraft, and it happened that Langley had been director of the Smithsonian up until his death in 1906, so the museums had direct access to his papers and prototypes. The Smithsonian changing position and conceding to the Wright brothers' version of events is more of a story of institutional memory rather than them making unreasonable demands

No, they're saying that the problems have been reviewed / play-tested by ≥2 humans, so they are not considered unfair or too ambiguous to solve in two attempts (a critique of some Arc-AGI-1 puzzles that o3 missed). They have a lot of puzzles so they were divided among some number of testers, but I don't think every tester had to try every problem.

It's a brand new website and old URLs won't work (this has been somewhat routine since Obama's first term). I wouldn't take that as a sign that a specific executive order is rescinded. However it may have been grouped in with other Biden tech executive orders (such as AI safety) which are being rescinded as excessive regulation

Chemistry applications, which may be possible using the current or near-future era of quantum computers. It helps to have actual quantum effects to compute with. For example (I am not affiliated with this company and don't know if it works) https://www.quantinuum.com/products-solutions/inquanto

Not really. The primary issue with Bitcoin x quantum computers is that older Bitcoin transactions contained a full public key (and not a fingerprint / wallet address), so a quantum computer could manipulate that to make a false signature on wallets including Satoshi's (this would be obvious and detectable if someone starts moving very dormant coins). For the larger network there are some other points where it becomes an issue, but before then you'd need to fix HTTPS web traffic and PGP, too.

https://www.deloitte.com/nl/en/services/risk-advisory/perspe...

I don't have any inside knowledge to expand on this, but it's interesting that Voting Village has had a public participatory component (website, social media, annual appearance at DEFCON, CFPs guiding messaging on topics on hardware and misinformation); the DARPA AIxCC appears to follow that model, and yet this Franklin Project is much harder to link to a website, social media, mailing list, etc. Is it too early, is it invite-only...?

It's likely an issue with displaying Arabic text in WebGL, since that needs to handle all of the possible options for each letter in the text-shaping.

For people unfamiliar, Semantic Scholar is run by the Allen Institute and has been researching accurate AI summarization and semantic search for years. Also they have support for author name changes.

There are some authors in common with a more recent paper "Defending Large Language Models against Jailbreak Attacks via Semantic Smoothing" https://arxiv.org/abs/2402.16192

An AI prompted to say it's an archivist might not have the same legal weight as one saying it's a doctor or can identify toxic mushrooms, but the organization suffers by bringing in tech which doesn't help the employees - especially when a big part of their job is finding reliable, cite-able sources of information.

I'm reading a book by a papyrus expert (Stolen Fragments) and I was surprised by how they frame it as safeguarding history first, and the community accepting or rejecting lines of work based on that. If you're an archivist you are aware that AI is part of the future, but also you're planning to still be an archivist at the institution and want to shape what that looks like.

From other reports, it sounds like hotel security was trained about hackers carrying the Flipper Zero. Maybe they thought the badges were something like that?

Video is from 2016, so tough to say how much of that is or ever was implemented. The NYC subway is just now piloting metal detector type devices which are poles which detect "signatures" of common items and weapons. I have no idea how often these detect weapons or false positive on a laptop or lacrosse stick, and the press has not been interested in the science of it so far.

My nomination for night sky viewing: Ölgii in western Mongolia (was there for the golden eagle festival). Clear desert sky, accessible by airplane, not a tiny town either.

I watched a "reaction video" to The Truman Show recently where the hosts were wholly unfamiliar with the movie beyond Jim Carrey. They focused on the cruelty of faking the family, relationship, and community around him. The film also shows that acting wasn't enough, and the corporation/creator essentially "cheat" to keep him from leaving and being his own person. So you can see that part of the movie without thinking it's no big deal in social media world.

I also watched a channel which uploaded daily vlogs during 2020, and the person said that as a kid she wondered if The Truman Show could be real and happening to her. And now she sort of created it for herself? Something to think about.

I don't think anyone has a solid public answer about why they didn't reopen with other museums after 2020. It might be too costly, people leaving the organization, etc. to the point that the board didn't think it could reopen.

Agreed, I gave real $ to the "Lambda Perpetual Access Fund" and now feel negative about the whole thing, but the sections where the author is talking about Austen's golf swing, bots, having written a paragraph conscious of clickbait? Not really necessary.

Job loss for knowledge workers, especially affecting intro jobs (I'm thinking of frontend dev or paralegal work, which is how a lot of people get a foothold in these fields). What are reasonable degrees to take out a college loan for in 2024?

Militaries using AI checkpoints, guards, patrols, and loitering munitions (from a Humanitarian / rules of war perspective, not necessarily existential risk). Possibly to continue regimes where human soldiers would defect or differentiate civilians from armed opponents.

AI-reinforced media bubbles promoting alternative realities (most people blame this on social media, but from QAnonCasualties it seems like YouTube and Facebook recommendations hooked a lot of people before mainstream companies assessed whether they could be harmful). I think this is more core than "deepfakes" because people believe even fake Tweet screenshots if it backs the narrative.

Based on the page footer ("IECC ChurnWare") I believe this is a site by design to waste time for web crawlers and tools which try to get root access on every domain. The robots.txt looks like this: https://ulysses-antoine-kurtis.web.sp.am/robots.txt

I don't see how this does much to keep bad actors away from other domains, but I can see why they don't want to give up the game for OpenAI to stop crawling.

These are the models on HF which have "Matryoshka" in the name, such as the PubMed one; I worked on a protein sequence one https://huggingface.co/models?search=matryoshka

Major kudos to this library for supporting Matryoshka/nested/adaptive embeddings (https://huggingface.co/blog/matryoshka), which I needed to train a model recently

What would prevent a known person from accepting a govt payout to sabotage their project, or to merge a plausible-looking patch? Relying on identity just promotes a type of culture of reputation over code review.

But how far away are we from hardware for stable quantum key generation/storage that would fit on a tabletop, much less a home laptop or smartphone? Almost certain that consumer devices will stay in classical computing and use PQE.

When I hear that this military base on US soil, with dedicated security and surveillance, can't recover one drone, or get photos to trace who or what sent the drone, I think of the London airport shutdown from 2018. There is no super drone swarm. https://en.wikipedia.org/wiki/Gatwick_Airport_drone_incident