trobertson's recent activity
-
Comment on It breaks a village in ~comp
-
It breaks a village
24 votes -
Comment on Overview of the proofs behind the Navier-Stokes $1 million millennium prize problem in ~science
trobertson Link ParentBy design, they are the first type. The Universal approximation theorem, which is the foundational theory in ML/LLMs, is all about universal approximation without care for subject matter knowledge...are these AI solutions more like the prospective prof's, where they just threw more numbers on the board until everything fit expectations regardless of the underlying physics, or our prof, where they actually model the behavior based on principles?
By design, they are the first type. The Universal approximation theorem, which is the foundational theory in ML/LLMs, is all about universal approximation without care for subject matter knowledge or understanding. Quoting wikipedia:
the universal approximation theorems (UATs) state that neural networks with a certain structure can, in principle, approximate any continuous function to any desired degree of accuracy.
Universality is achieved by increasing the number of neurons in the hidden layer, making the network "wider." Other versions of the theorem show that universality can also be achieved by keeping the network's width fixed but increasing its number of layers, making it "deeper."
a.k.a., just chuck more variables at it.
-
Comment on What games have you been playing, and what's your opinion on them? in ~games
trobertson LinkI have just stopped playing Palworld. I think I'd give it a 6/10; it was a lot of fun in the early hours (9-10/10), but the further I progressed the worse it got. Towards the late game it was just...I have just stopped playing Palworld. I think I'd give it a 6/10; it was a lot of fun in the early hours (9-10/10), but the further I progressed the worse it got. Towards the late game it was just not fun at all (2/10) and the charm of it had worn off, leaving me with no interest in continuing the game.
But the early game is really fun so I made a few different worlds, each with their own settings, to try to extend the fun through to the whole game. Even going so far as essentially cheating through the bad mechanics (crafting + resource farming) with both world settings and mods. In each of those playthroughs I dropped out at about the same point: the need for chromium and coralite.
This is the point where the game is just a grind. Up to this point you can make work stations to passively produce resources. chromium and coralite need to be manually farmed, by you, with no direction provided as to where to find it. Reddit + wikis tell you this, but the manual nature of acquisition turns the game into nothing but resource grind. All of the fun parts go away and you are left with rote monotony. Maybe this is fine in multiplayer, where you can shoot the shit while doing crap labor, but in single player it is a run killer.
I have some other issues with the game, but next to the game-stopper grind those issues don't really matter.
(Except performance: I have a strong machine and had shit-tier performance, and crashes, all throughout my playthroughs. Based on conversations I found while looking for solutions, it appears to be essentially random as to whether a given machine will have good or bad performance.)
Verdict: Worth the twenty-some dollars I spent on it, but I'm done with it for a long, long time.
-
Comment on What games have you been playing, and what's your opinion on them? in ~games
trobertson Link ParentGiven that there is not a linear story, you are playing correctly. For me Outer Wilds got better and better as I progressed, peaking at the end of the game. If you are at all intrigued by the...Given that there is not a linear story, you are playing correctly. For me Outer Wilds got better and better as I progressed, peaking at the end of the game. If you are at all intrigued by the stuff you have found then you're doing just fine.
-
Comment on GOG added OpenMW and Tamriel Rebuilt into their store for free as mods for TES3 Morrowind in ~games
trobertson Link ParentI readily attest to holding this belief. Narsis (the Grasping Fortune release) is the best Elder Scrolls, hands down. I almost can't believe it's real. And it's not the first example of...Plenty of their newer regions are considered to be better than the base game regions.
I readily attest to holding this belief. Narsis (the Grasping Fortune release) is the best Elder Scrolls, hands down. I almost can't believe it's real.
And it's not the first example of exceptional quality. The releases prior to Grasping Fortune added some Tribunal Temple questlines that showed an incredible understanding of the push and pull of politics on the operation of a religious institution. The various Temples of Bal Foyen, Almas Thirr, and Hlan Oek (all three of which are in the same general region) are each shaped by the forces they live within. These three questlines, considered together, far exceed the depth of understanding in just about any other Elder Scrolls content. And, generally, they are also good quests.
At this point I semi-seriously consider PTR (Project Tamriel + Tamriel Rebuilt) to be the real Elder Scrolls continuation. Bethesda's games feel like bad fanfic in comparison, and I have no hope whatsoever for ES6.
Also, OP, PTR is very active. The Poison Song release was just a couple days ago. It make Great House Indoril joinable and gives it a questline in addition to remaking the region the House occupies. I haven't played it yet, but I'll be doing so within the next couple months.
-
Comment on Bethesda, id Software reportedly hit hard by Microsoft layoffs in ~games
trobertson Link ParentEspecially so when the claim is that Bethesda will now focus on core franchises, which include... Doom.Especially so when the claim is that Bethesda will now focus on core franchises, which include... Doom.
-
Comment on Mullvad CEO gives ≈€452,000 to the Örebro Party in ~society
trobertson LinkIf you, like me, had no idea what the Örebro party is: https://en.wikipedia.org/wiki/%C3%96rebro_Party#Ideology I'm still a bit confused on what they are actually trying to accomplish, but I'm (in...If you, like me, had no idea what the Örebro party is:
https://en.wikipedia.org/wiki/%C3%96rebro_Party#Ideology
The ideas of Markus Allard has been described as national conservative and "conservative with a Marxist stamp".[27] Expo has described the Örebro Party as part of a growing left-wing nationalistic movement in Sweden.[28] The party has also been described as both right-wing populist[29] and left-wing populist as well as left-conservative.[30] Party leader Markus Allard was referred to by the Waste Ombudsman at the Swedish Taxpayers' Association as a communist, nationalist and a populist. The chief editor at a local newspaper in Örebro made accusations against the Örebro Party and said that it was his personal opinion that they are "racists, fascists and nazis".[31]
I'm still a bit confused on what they are actually trying to accomplish, but I'm (in my head) putting them in the same bucket I put the "authoritarian left" groups into, as it seems they share a lot of concerning behaviors and views.
-
Comment on Bureaucracy is eating the world in ~society
trobertson Link ParentI couldn't help but scoff at those two paragraphs. I don't mean to be rude, but I will be rebutting quite a lot of the claims in the posted snippet. Disclaimer: I'm an American by birth. Do excuse...- Exemplary
I couldn't help but scoff at those two paragraphs. I don't mean to be rude, but I will be rebutting quite a lot of the claims in the posted snippet.
Disclaimer: I'm an American by birth. Do excuse me for defending Europe from this late-stage-capitalist slander.
In 2000, incomes in the original six members of the European Union were just 10 percent behind Americans. Today, they are 20 percent lower.
Mean? Median? Does it matter at all when looking instead at quality of life? Incomes are only useful in-so-far as they provide you a quality of life; the raw number mostly doesn't matter. The life the income can provide you is what matters.
I'd argue that most Americans, under Trump's idiocy, are generally worse off than most Europeans. Higher incomes don't offset nation-wide insanity.
One factor behind this has been the lack of innovation in European business.
This is a claim that gets repeated frequently, but what the fuck does "innovation" even mean nowadays? Does something like Ozempic count? Or is it only
"tech"ad-tech companies? Chatbot companies? Does "innovation" mean "stock market goes up"? If it does, I grant you that the USA has manipulated the stock market to hell and back, what with reintroducing stock buybacks (which used to be illegal as they are blatant stock price manipulation).Tesla is now worth more than the next nine largest carmakers in the world put together.
Tesla is a meme stock. Pretending that it is a serious economic indicator only serves to identify the author as a joke. Observing Tesla's car business, they are in a total global collapse. Their competitors are mostly doing okay (but are legislatively insulated from BYD, which is blowing everyone out of the water, including Tesla).
Six American cities are now served by robotaxis made by Waymo.
95%+ of European cities are served by good public transit, which does a much better job than Waymo at a much lower per-trip cost. Pretending that cars are a good way to traverse cities is pure nonsense.
Understanding why it doesn’t have a Tesla is existential.
Meme stocks are not existential.
And just seven companies globally – Google, Apple, Amazon, Meta, Microsoft, Samsung, and Huawei – spend more on research each year than Volkswagen.
No shit? Gas/diesel cars are a solved problem. Volkswagen's primary need for research is the move to electric vehicles (which they are doing), and their limit is their available capital. Cars are not a high margin business; Volkswagen needs to be much more careful, per dollar, than the high-margin ad-tech companies do.
Also, "research spend" is a very nebulous term; bog standard software development is considered R&D by most jurisdictions. So it makes sense that companies that make their money via software development are also the ones who have a big number in the "R&D" box.
-
Comment on The US Supreme Court's Republican appointees end civil rights redistricting protections in ~society
trobertson Link ParentRBG was a big part of the court's decline; she could have resigned under a Dem and been replaced by someone reasonable. Instead, she held on to her seat until she died (under Trump) and basically...RBG was a big part of the court's decline; she could have resigned under a Dem and been replaced by someone reasonable. Instead, she held on to her seat until she died (under Trump) and basically gave her seat to the far right.
Aside from that enormous blunder, her record was pretty good.
-
Comment on Native American Indian porcupine quillwork, how to do porky quillwork in ~arts
trobertson Link ParentKagi is a paid search engine, so if you can't load the linked image results, here are some pages on the artist with a few examples of her work....Yvonne Walker-Keshick, who does [amazing, painterly quillwork] <- Kagi link
Kagi is a paid search engine, so if you can't load the linked image results, here are some pages on the artist with a few examples of her work.
-
Comment on What games have you been playing, and what's your opinion on them? in ~games
trobertson Link ParentIf you haven't already, check out Lunacid. I put ~30 hours into that a few weeks ago and it's a great KF-like.If you haven't already, check out Lunacid. I put ~30 hours into that a few weeks ago and it's a great KF-like.
-
Comment on Best of Humble Bundle: Beamdog & Owlcat: RPG Masters (pay what you want and help charity) in ~games
trobertson Link ParentI can endorse both of the Pathfinder RPGs, Kingmaker and Wrath of the Righteous. They are both great, and if you've played the tabletop adventure module, you'll roughly know what to expect....I can endorse both of the Pathfinder RPGs, Kingmaker and Wrath of the Righteous. They are both great, and if you've played the tabletop adventure module, you'll roughly know what to expect.
Kingmaker is game where you and your adventuring party build a kingdom in a hostile land. Kingdom management is a large part of the experience, for better or for worse. The mechanic is divisive but if you like cRPGs this is definitely worth trying.
WotR is a war campaign against demons. It is a much better realized vision (being the second game by the team, after Kingmaker). Characters are better, there's more mechanical depth and story depth, etc. But the war campaign mechanics are universally considered to be a cheap rip off of Might and Magic. There's a mod to auto-win these battles, and I suggest you get it.
Both games have the best character creation and build planning in any cRPG ever. For many people the character creation screen is the game (in a good way). Coming up with new ideas for characters is incredibly addictive. This is in direct contrast to something like BG3, where it almost felt like the builds were on rails. BG3 gave you maybe 1 choice, on average, per character level; KM and WotR give you on average 3-5 choices per character level depending on class.
-
Comment on <deleted topic> in ~enviro
trobertson Link ParentMy bad, I didn't see this was already submitted. I'll remove this one.My bad, I didn't see this was already submitted. I'll remove this one.
-
Comment on <deleted topic> in ~enviro
trobertson LinkTechnology Connections gives a breakdown of why renewable energy makes sense in comparison to fossil fuels. This video is not for people who are already sold on the transition to renewables. It is...Technology Connections gives a breakdown of why renewable energy makes sense in comparison to fossil fuels. This video is not for people who are already sold on the transition to renewables. It is aimed at people who have been made skeptical through anti-renewable propaganda.
-
Comment on The 2025 Steam Awards in ~games
trobertson Link(Sorry, I did a essay.) Were you to ask TGA, they would (not actually, of course) say that there was only one game released this year: Claire Obscur: Expedition 33 (E33). That game swept TGA in a...(Sorry, I did a essay.)
Were you to ask TGA, they would (not actually, of course) say that there was only one game released this year: Claire Obscur: Expedition 33 (E33). That game swept TGA in a way that no other game ever has.
Whereas Steam's awards are a much wider spread. Across 11 categories, only a single game won two awards. Steam's awards demonstrate the width and breadth of gaming. TGA did not.
It comes down to the difference in voting populations, of course. TGA is game journalists, Steam is (PC) players. And this year shows the gap between the two.
Let's first look at E33, as that really is the sticking point for a lot of people.
TGA awards for E33
- Game of the Year
- Best Game Direction
- Best Narrative
- Best Art Direction
- Best Score and Music
- Best Performance (Actor)
- Best Indie
- Best Indie Debut
- Best RPG
Aside: In TGA, E33 was also a finalist for Best Audio Design.
Steam Awards for E33
- Best Soundtrack
Aside: In Steam's awards, E33 was also a finalist for Game of the Year.
It's clear that journalists were taken with E33. They loved it.
Or, being cynical: they loved what they saw of it. Games journalists have a well-earned reputation of only playing a game for a few hours before writing their reviews and moving on to the next game. And I suspect that's a major contributing factor here.
E33 has an absurdly good first several hours. The characterization is great, the narrative is great, the visual are great, the music is great, the direction and presentation are great. It has heartfelt character moments with characters we've only known for minutes. It present as a gut wrenching story of a dwindling population coming together to overcome adversity.
For those who don't finish games, those first hours make E33 seem like the best game there ever was. And if that's all you play, then I can understand how you'd give it all of the awards at TGA.
But beyond those first hours things change. There are some wild (WILD) tone shifts in Act 1 that persist throughout the rest of the game. There are major story beats and revelations that sabotage a fan-favorite character, and recontextualize the rest of the characters... for the worse. The story is not what was presented in Act 1.
Personally, I would call the story (of Act 1) a bait-and-switch. The real story only shows itself in Act 2.
So I think it makes sense that gamers at large (via the Steam awards) did not rate it as highly as the journalist voters of TGA. E33 is still a beloved game, but it is a game with flaws. It does not feel right for it to sweep the awards. Especially not when the year was so stacked with excellent games.
I considered writing up a list of the many excellent games this year... but I don't have to. Look at the Steam Awards list: across 11 categories, there is only one repeat. If you remove the "Game of the Year" all-up category then there are no repeats.
And that feels much more correct. There were tons of amazing games this year. Many more than can fit in those 11 categories; in total, there were 50 different games nominated for those 11 spots in the Steam awards.
So... TGA is trash and Steam is great? We've all known that forever. But TGA is a marketing show, and Steam's awards are a glorified blogpost. So guess which award set gets all the attention, and consequentially, the credibility.
-
Mathematically extra-complicated Secretest Santa 2025
14 votes -
Comment on Fifty Shades of OOP in ~comp
trobertson Link Parent@Akir @tauon I don't mind giving you an invite, but you'll need to private message me an email address. This is what the invite process looks like: https://imgur.com/a/5TkxXBKI don't mind giving you an invite, but you'll need to private message me an email address. This is what the invite process looks like: https://imgur.com/a/5TkxXBK
-
Comment on Fifty Shades of OOP in ~comp
trobertson Link ParentProbably, yeah. Ever since HN degraded* I've been going to lobste.rs/active for programming discussion. Actual practitioners and academics talking shop, divorced from big-corp motives and...Should I be on lobste.rs?
Probably, yeah. Ever since HN degraded* I've been going to lobste.rs/active for programming discussion. Actual practitioners and academics talking shop, divorced from big-corp motives and bullshit.
* HN users/posts nowadays are 30% fascist, 50% hype/shill/ads, 10% worthwhile, and 10% oblivious to just how bad it's become. dang, in particular, is the poster child for the latter group.
-
Meta is earning a fortune on a deluge of fraudulent ads, documents show
53 votes
An observation of how a change in AI policy has affected the Bevy game engine community in the months since a new AI policy came into effect.
As can be inferred from the title, the author, a long-time Bevy contributor, has had a very negative experience with the changes in the community. Their post describes the new policy's effects on the behavior and health of the development community, and how the changes in the community have directly decreased the author's willingness to continue participating in the Bevy project.