er4hn
er4hn on Hacker News
There were a number of sea lions hanging out there! It's entirely possible they were snacking on whatever salmon were going by. Not all was lost though, I got to enjoy watching some boats go through the locks.
I'm assuming it was out of season because in talking with the Uber driver afterwards he said we'd missed the end of the season. I'm assuming his "local model" was up to date and accurate.
It's hard to try and compare these benchmarks against real world use cases. As an experiment I let Meta Muse make all major decisions on a weekend trip to Seattle ~2 weeks ago. I think this would have been using Muse Spark 1.3, which is equivalent to GPT 6 - Sol.
On the plus side it told me taking the light rail from SEA into the University district was faster and cheaper than Uber.
On the downside it would recommend restaurants that were out of business, checking out Salmon swimming through locks out of season, got the opening of Beecher's Cheese in Pikes Place wrong, and fumble it's way through most simple planning decisions. I think my favorite was when it asked me what flight I have tomorrow so it could help plan when to head to SEA. It sent me a very panicked message that the flight was leaving today! Very soon! I had to remind it this is a daily flight and I am booked on the next day before it checked and apologized.
For things that are verifiable, and code tends to be that way, these models can do all sorts of things. But I find it hard to reconcile Meta's goal of having a personal assistant bot with something that is unable to do the simplest tasks.
I get this unless I drink a lot of water. About 48 fl oz seems to be the right amount for 10g creatine. Sometimes this needs to be combined with a light snack, like a handful of nuts.
Same here, the quote on time passing stuck with me quite a bit.
I was given a picture cube, which is like a Rubik's cube but every side is a unique picture. It came scrambled and I don't have an original reference image. I like to take videos of it and give it to llms to solve. I call it my agi test because it hasn't been solved yet
very fun game. I don't live there any longer but I "ran" the commute from my apartment to work. It felt very nostalgic to go through this.
The purchasers of the cameras, ie HOAs, law enforcement, etc, own the data. They are also the ones routinely caught abusing this. This is a real problem that should be dealt with by enforcing laws against the people improperly using the data.
I'm not sure what a realistic solution is for Flock to try and manage data they do not own nor if it makes sense for them to deny access to data they are not the owners of.
signull is more of an anonymous sh*tposter than a known industry insider, but I think this does capture the sama contribution to OpenAI very well. At least from an outsider who follows this stuff based on vibes.
This page came about because of how long it took PQC to get standardized. This was a slow enough process that a whole slew of QKD vendors arose and sold a lot of products promising this as a solution to dealing with quantum computers and harvest now decrypt later attacks. Many of those products did not do a great job at actually preventing listening in on their lines since QKD is an ongoing field of research where new issues are routinely being discovered.
This must be what gwern meant when he said to write for AI.
Right, that's the same situation as fuzz testing today, which is why I compared it. I feel like you're gesturing towards "Attackers only need to get lucky once, defenders need to do a good job everytime" but a lot of the times when you apply techniques like fuzz testing it doesn't take a lot of effort to get good coverage. I suspect a similar situation will play out with LLM assisted attack generation. For higher value targets based on OSS, there's projects like Google Big Sleep to bring enhanced resources.
I think the author makes some interesting points, but I'm not that worried about this. These tools feel symmetric for defenders to use as well. There's an easy to see path that involves running "LLM Red Teams" in CI before merging code or major releases. The fact that it's a somewhat time expensive (I'm ignoring cost here on purpose) test makes it feel similar to fuzzing for where it would fit in a pipeline. New tools, new threats, new solutions.
It's clear, from watching Russia fail to be completely sanctioned that this is not watertight. The question I have is: have these sanctions added a money laundering tax to doing business? How much? What is the cost of enforcing the sanctions vs the added cost and is that worth it?
I don't know if this has been explored, bit I think it's an interesting follow on to "all or nothing" watertight sanctions.
In theory it prevents failures of the allocator that would allow reading uninitialized memory, which isn't really a thing in Go.
In practice it provides a straightforward path to complying with government crypto certification requirements like FIPS 140 that were written with languages in mind where this is an issue.
Speaking (unofficially) as someone who works at one of the "other brands" that reeks of journalists having a bias.
Well the idea behind tokens is that they should be time and authZ limited. In most cases they are not so they degrade to a glorified static password.
Solutions like generating them live with a short lifetime, using solutions like oauth w/ proper scopes, biscuits that limit what they can do in detail, etc, all exist and are rarely used.
Not having redundant rails in case of breakdowns is something BART is well known for
How many files do you have? At what scale did you see this being a problem?
I'm a fan of Obsidian, not affiliated with them, but my experience with basic file syncing like syncthing or git is that you should be able to easily get up into the ten's of thousands of files without an issue.
Wandering around my dungeon, doing things, in the first person was such a fun experience as well. It felt like such a novel way to explore what I was doing and look at my creatures "in the face" so to say. In some ways it feels like a precursor to things like minecraft, where you could do some tasks as an imp in the dungeon.
Is this the financial version of the 3-body problem?
There's a long history of this. A defense against the Morris worm made use of this as well.
> Why would we want a digital currency? For similar reasons to all the other stuff above. It's more convenient. When you "transfer money" from your bank account to another, your bank has to physically move the associated cash from it's vault to the other banks vault, by hiring secure trucks, people, and so on. If the money has to cross a border, that's even more of a hassle, now you have to physically cross a border with a truck full of cash. When a bank "holds onto your money", they need a big vault full of cash, they have to count it, account for every dollar, physically safeguard it, etc.
>
> This is a huge cost, inefficiency, and a big challenge of banking, and it's one reason transaction fees and banking fees are so high.
That's absolutely not how this works though. Banks perform electronic transfers and most of the money is accounted for in databases. The problems are slow, antiquated, technology, which is made worse by the amount of regulation surrounding it that makes it hard for new contenders to enter and drive down prices via competition.
Cryptocurrency is trustless, but there is an interesting tangent about if you _do not_ want a government to control monetary policy.
No, it's so much worse than anyone could have imagined. For example: "kids’ shorts with drawstrings longer than regulation length, which cause a trip hazard."
This is really fun, thanks for putting it together. I appreciated the checksum and entropy visualizers quite a bit, they made how it works "click" for me.
That was something I struggled to understand for AI-2027. They have China nationalize DeepCent so there's only one Chinese lab. I don't understand why OpenBrain doesn't form multiple competing labs because that seems to be what happened IRL before this was written.
It is, but the collaboration portion is a CYOA part you need to implement yourself. There are OSS versions of that as well but they are not officially supported.