a_vanderbilt

I found the AI 2027 paper to be overly optimistic, but not wholly fantastical. This paper feels wildly speculative, and relies on premises I am not confident even pass surface reasoning. Even under optimistic conditions, we are not going to see robots "capable of 95% of all cognitive and physical tasks" by 2035. Nor do I think a 74% unemployment rate is even remotely possible. Economic collapse would implode AI development long before those figures were plausible.

I'm really on the fence about the Steam Machine. The Deck had the benefit of occupying a market niche no one else figured out yet, and the price was good. With component prices at an all-time high I'm not seeing how they can get the price at a compelling point. The living room niche is already occupied by consoles, and that's the first thing people are going to compare it to.

For many, the appeal of a Mac is that it isn't running Windows. I'm not seeing how this won't be a repeat of the OG ARM Surface, just with a higher spec'd GPU.

I used to help build the CTFs for BSides Orlando. I ended up moving to another con, and at our last event we collected extensive logging for post mortem analysis.

We found that AI usage is basically guaranteed now, but certain challenge designs did thwart it. Challenges built with temporal visual elements made AI fall flat on its face, as it could not ingest/process the data fast enough to act on them in time. We also found that counterfactual challenges (ie. the result you get did not match what we suggested you'd get) made AI-assisted solve time slower compared to pure humans, indirectly penalizing over-reliance on AI. Multimodal challenges combining audio and visual elements were also very effective, but were not as accessible to players.

This paper gave us some ideas about designing those challenges: https://arxiv.org/pdf/2308.02950.

For our next event we figured out a way to thwart AI in our CTF: embed the CTF in a game engine. The loop essentially becomes something like this: Connect to a simulated access point in the game, the K8s cluster connects their attack container to a private network with the challenge box(es). Hacking the boxes doesn't render a flag, but rather changes in game state. AI did very poorly coping with this in our testing, as it can't derive the spatial state of the game world very well and it soft decouples the inductive reasoning loop it relies on to know if it is on the right track.

The downside to this approach is it is far more labor intensive for CTF organizers, and requires players to have a computer capable of running the game. We are also betting on AI to not advance enough by the time we ship to be able to just ingest the entire game state in realtime and close the loop that way.

You should not consider Tim Sweeney's comments on the matter as a reliable source. He was veiling his true motivations behind that statement. The Switch does not run Linux either, it's a custom OS descending from the Wii's iOS.

The cheating issue isn't really a matter of being able to run custom kernel code. You can do the same thing on Windows, which is why remote attestation is a thing for some games. As someone who has developed games for Linux (and Windows / Mac), it's an endless cat and mouse game. So long as the system can execute code that is not yours, you never really are getting perfect anticheat. Ease of loading custom kernel code isn't really a hurdle to that.

I find that client and server based in combination is the robust approach. I once implemented anti-cheat in which the server lied about game state, which a regular client without cheats would act predictably on. Deviation from that behavior is a useful heuristic to build a suspicion score.

That's the same mindset some people had 3 years ago when they said AI wouldn't be capable of software development. Look where we are now.

In theory, the market should be pricing in based on future potential. As it has become increasingly clear this past decade, the market is not rational.

Sonarr et al.

My experience was the same while helping to adapt a Steam Deck game for wider Linux support. The issue wasn't Waylandisms, most of those have already by figured out. It was GNOME. Their preferred resolution to issues seems to be dropping support rather than bug fixes, and they go out of their way to adopt implementations that are against the momentum of the wider community. I can get why they make some of their decisions, but things like killing the tray indicator or server side decorations are insane. To be an outlier in name of a greater or grander goal is one thing, then there is whatever GNOME is doing.

It might be wordsmithing to skirt around "robot" as a fully autonomous entity. Much like their FSD, I expect they aren't going to deliver full autonomy anytime soon.

So much of it is a problem of execution. If people could use Linux without ever having to know what a terminal is (much like the average Windows user doesn't know what PowerShell is), then it would actually be quite successful. It has gotten better over the past decade, but it still suffers from endless paper cuts and the odd issue that requires a shell session to fix. I will say that Valve's SteamOS has come the closest to avoiding this trap. You can use a deck without ever having to touch a CLI.

It's been an unfortunate re-occurring issue for me as well. Recent hardware is much better about this, and I too have seen the performance bumps at the cost of software compatibility. I feel like if Adobe brought their CC suite to Linux I'd have no reason to ever use Windows outside the random game that _needs_ it.

At least on the latest Sequoia, there has been no hard requirement for an online account. They nudge you towards it, but you can decline and continue. As far as I can remember, macOS has never required an online account to set up a Mac.

Hello back pain my old friend. Can you imagine the spinal G forces of getting yoinked by an airplane even at slow speeds?

Atlas being created is kinda the shot across the bow. You can integrate with us willingly, or we'll hook into your web apps anyways. One retains at least some control. Same outcome as Disney's deal with Sora.

LadyBird cannot come fast enough. I'm not being dramatic when I claim that this will be looked back upon as the nail in FF's coffin.

To me it would mean that a git repo should not have scripts, runners, etc. configured that we don't have the means to easily and readily replace. It should all be documented and understood well enough that we could kill the repo and init another at will.

A core SRE principle is that "machines/servers are cattle, not pets". They shouldn't be special or bespoke in a way that makes replacement painful or difficult.

Essentially correct. With a quantum computer you do multiple runs and average the result.

And Blink. I used to contract with them a few years back, they all rely heavily on open source.

This approach is either a means to delay and allow them to hedge, or it's a shortsighted attempt to stop the inevitable. When Google ships their competitor to Atlas, the outcome will already be decided.

The US congress is partially to blame. They saw Japan's rising influence in the digital/industrial space and threatened them with sanctions. See https://news.cgtn.com/news/2019-07-11/Lessons-from-U-S-Japan... for Chinese state media's interpretation.

The definition of AGI is diffuse enough to make it an argued point - until we can mostly agree it's already happened. For now, the stats are improving well enough across the industry to maintain investor attention. Will it all come crashing down a-la the .com bubble? It's seeming more likely by the quarter.

Like the digital economy post .com burst, I think AI will survive and grow far beyond its current market of chat bots and agents. The weakest will die, but the market will be better off for it in the long run.

The next big problem for AI is time horizons. Frontier AI has roughly doctorate level knowledge across many domains, but it needs to be able to stay on task well/long enough to apply it without a human hand holding it. People are going to have to get used to feeding the AI detailed and accurate plans just like humans, unless we can leverage an expanded form of leading questions like GPT-5 does before executing "deep research". Anthropic feels best positioned to do this on a technical level, but I feel OpenAI will beat them on the product level. I am confident that enough data can be amassed to push time horizons at least in coding, which itself will unlock more capability outside that domain.

I feel it's very different from Tesla, because while Tesla barely ever got closer to their promises the AI industry is at least making visible progress.

Unfortunately, loss of buyers often results in them just leaving the crop to rot, as moving it costs more than what they'd make selling it. Not that they should be surprised though. From soybeans to AI silicon, China has shown it is willing to take the hit and source alternatives to US products.

Given the advancements of EVs, the death of the muscle car was inevitable. They are almost universally faster, more performant, more efficient, and more "cool" in their designs. Do gear heads dislike them? Sure, but they aren't going to be the market forever. As energy costs rise, natural market forces will shift the demands of the buyer.

Culturally, the Japanese aren't likely to care. Take a look at Linux usage in Japan to get what I mean. You will have a small but very dedicated group of users who won't change for anything, and then the masses who just use what is convenient. They don't like tweaking.

Absolutely would not be a good candidate for a gov service in the US. Imagine the politicization currently applied to the Postal Service, but on the scale of everyone's email?

It's insane numbers like that that give me some concern for a bubble. Not because AI hits some dead end, but due to a plateau that shifts from aggressive investment to passive-but-steady improvement.

At least in OpenAI's case, it raises the bar for potential competition while also implying that what they have behind the scenes is far better.

Most people don't really care all that much about the distinction. It comes across to them as linguistic pedantry and they downvote it to show they don't want to hear/read it.