Andrei Chiffa

Ok people, you were all right and I guess I got a bit carried away with the true scale of the drama and owe you an apology.

Most of us who spotted them left comments under videos, which sometimes were buried, sometimes rose to the top. That's a proper journalistic investigation - and you might be giving GN ideas if he decides to keep on going. For now it's a highly likely mechanism chain with anecdotal supply. His takes on Apple's hardware ecosystem closedness without ever accounting for the software side of the things (8 years of patches and support vs ~0, maybe 2 if you are lucky), or the market fragmentation making it impossible to find parts (replaced two batteries in iPhones and Apple watch, could not find parts for any Android my parents owned after even 2 years), definitely led to perverse incentives for both Android phone manufacturers and led consumers buy phones that became e-waste 2-3 times faster than if they were "mindless Apple sheep".

Vice and opportunistic troll has. The problem with Linus/GN/Madison is that all three of them has a lot of hardcore fans that have nothing else going in their life and value integrity and being generally nice to people above all else (comes with having been a bullied nerd kid I guess). They are not stopping, ever, and there are enough of them to keep downvoting his videos and squatting comments/reddit/twitter/... until things are set right. The problem with an extremely loyal fanbase, is that when they turn on you, they will hound you.

Tbh, he should have found a C<X>O for each of his sub-branches and trusted them to do a good job instead of personally reviewing every single draft and video before release. He did it for Mac Address, and it was arguably the best decision he ever made wrt covering Apple products, because his own takes on anything Apple-related are completely clueless. Not that his takes on other topics were highly informed (Twitter, RTX 4090 pricing, unionization, ...). Had he the cover of the other C<X>O, he would have had the excuse of "that's my opinion, but realistically <name> would not allow me to do it if it's not a good idea, in the same way he did it with Yvonne on a WAN show when it came to the Creators Warehouse and Labs investments. I have no idea if Terren Tong was the right person to become the CEO of the LMG, but someone like him was loooooong overdue.

No, but this has no importance on the internet celebrity wars. Every single of LTTs videos will be downvoted into oblivion and brands sponsoring them - receive a ton of hate, until something drastic is done about the situation. Rabid Linus vs reasonable Luke is a running joke, but realistically something just as drastic as a swap might be needed in the short term for a full recovery.

They weren't expecting to have to deal with Madison going public or leak of internal recordings confirming it. So one week was only to deal with GN allegation and damage control over Linus' response on forums here. Madison's allegation ins themselves were pretty bad, but the additional sex joke innuendos in the excuse video managed to not only add fuel to the fire, but made a number of long-term LTT supporters and Linus fans to pull out their pitchforks. At this point they have to first address this, or they will have all their videos downvoted into oblivion and no one left willing to sponsor their videos in the future. And that probably will need more time and something more drastic than a single week moratorium. Probably something like Linus going on an executive leave and Luke taking over the CVO role in the interim for at least a couple of months.

Had the same question yesterday, so let me re-link again the excellent @McGherkin's summary

Tbh I don't think that such an environment should be accepted in any workplace (but I am also a European), but IMHO the problem of the LMG is that they build their image specifically as the type of people who would not be doing it. Given the amount of shit Linus gave to other companies that did not do their work properly (starting with cancelling tours and booth visits stories on WAN all the way to Anker), fans listened to LTT's advice and used their guidelines to drive major purchases. GN's video basically suggests that in some areas they don't clear the bar they set for other companies they set, and Linus' initial response pretty much confirmed all of it, which understandably made a lot of fans with a Manichean vision of the world take LTT off the "Good little guy" list and put them on the "Evil large corporation" one. So basically GN video by itself could have been a "woopsie, thanks Steve for expressing your concerns and let us look into it and get back to you", followed up with a "yeah, we fucked up with the waterblock and the mouse; here is why it happened, here is how we are ensuring it will never happen again, and how we are making it up to companies that got hurt in the process", leading to a teary redemption and even more Linus fandom; but instead became a spiraling disaster that cannot be blamed on anyone but Linus because of his knee-jerk reaction. However at this stage it is much more about Madison, because LTT does have a reputation for a bro culture and when Madison left, Max (only other strong character young female public-facing employee before her) suggested that she left for the same reasons as Madison. As far as LTT reputation goes, it's hard to do worse.

Had the same question @McGherkin did an excellent write-up in response:

In other news, Reddit top page is 8 posts about LTT out of top 10. Same ratio going down into the top 40. I guess LTT finally made it on other platforms, but I doubt they will have a lot of sponsors remaining after this incident.

It's an excellent summary with some great sense of humor - thanks! Overall, to me it looks like LTT is going down the way of TechCrunch and Linus is going down the way of Michael Arrington, for exactly the same reasons. I do hope that they'll be able to stop, but I did leave my fair share of comments going in the same direction as GN under Linus' videos over the last year and am not surprised by the allegations from Madison (Max made the same when she quit). Years ago (as in 2015-2016) Linus made a comment on WAN show about the "deplorables" being his core audience. I don't think he is wrong - in that tech is probably a way to avoid them gravitating towards extreme right and manosphere - but that also probably needs to be articulated to female employees who will attract their attention, as well as fans that graduated to employees from that community; and made excessively clear internally on the HR-policies-that-cannot-be-overridden-by-Linus upfront for everyone joining.

Ok, for those of us living under a rock, what exactly happened that needed that video? Like jokes have been increasingly tasteless and some takes - rather dumb over the last year or two - but what triggered the release of that video specifically? And why it had no context to it?

Thanks, but I know my model number and have already ordered replacement for it. What I am interested in now are reputable manufacturers who would have batteries that would also last 3.5 years instead of 1.5.

I have a lovely Dell Inspiron 15 Gaming that is still going strong after > 5 years. The only part that truly gave up after all that time is the battery, which I replaced back in late 2021, after 3.5 years of use, after it could not hold an hour of document editing work. The replacement one I got then (K KUYER) breathed a new life into it, but it is now getting to the same threshold after only 1.5-ish years, looking that the replacement part was not as good as the original. Is there anywhere a list of quality replacement batteries manufactures, to know which batteries would last; to minimize e-waste generation?

Thanks for taking the time to respond, Riley! While I absolutely agree that no corporation is to be "trusted" on their word (and that's why I am looking forwards to more data coming out of LTT labs), some companies do a better job at aligning their actions with their statements. Some companies do extremely well. Currently, among large corporations, it's Hugging Face putting their money where their mouth is; but before Musk Twitter AI/ML safety team, for all the flack they were getting, were doing a pretty good job and not only letting external researchers look at their data for free, but also acting on the reports they were receiving. While OpenAI has it's issues, they do have team that works on negative impacts of LLMs and that has impact into what goes into production. For all the jokes the "we are not releasing GPT-2 due to the risk of misinformation" got in 2019, retrospectively, they were not wrong. This continues to this day - GPT-4 paper contained 50 pages of description of critical failure modes and what they did to mitigate it; and they are accepting external collaborators who want to break their products and publish their findings, after OpenAI patches systems in production. Unfortunately Google until now has shown a callous disregard for ethics and safety of AI/ML. Google search top blurb has been broken for years in dangerous ways, and nothing was done to mitigate even widely reported errors. YouTube algorithm has been hijacked without them noticing or doing anything about it until it grew into a major scandal with legal ramifications (ElsaGate/DistrubedElsa). If we focus just on recent LLM developments, the LaMDA paper (which is the architecture behind BARD) only has 1.5 pages on safety, with 0 evaluation examples or results provided. Sparrow - the next iteration of conversational search LLM Verge recently confirmed they were working - does barely better, by bumping it to 3.5 pages and showing some experimental results. To the best of my knowledge, no one external is let anywhere near their production models or algorithms, nor do they act on reports based on user interfaces (but I think you noticed it already with spam in Youtube comments).

Thanks for the great video Riley! However, a quick note about Google and AI safety. Every time they talk about their concerns about AI safety and ethics, it is important to keep in mind that they consistently sacked their ethics and safety teams over the last 2.5 years, pretty much removing anyone who would be doing actual work to make AI safer and more ethical. Here are just a couple of major stories about it. April 7th 2023, New York Times, Google suppressing an AI safety paper and pushing researchers writing it out: https://www.nytimes.com/2023/04/07/technology/ai-chatbots-google-microsoft.html May 2nd, 2022: New York Times, Google fires researchers who challenge a paper from its own research team hyping AI performance: https://www.nytimes.com/2022/05/02/technology/google-fires-ai-researchers.html Feb 19th, 2021: The Verge: Google fires the head of its AI safety team due to their stance on LLMs: https://www.theverge.com/2021/2/19/22292011/google-second-ethical-ai-researcher-fired December 4th 2020: MIT Tech Review: Google fires one of its top ethics and safety researchers over a paper warning about the danger of LLMs (Stochastic Parrots): https://www.technologyreview.com/2020/12/04/1013294/google-ai-ethics-research-paper-forced-out-timnit-gebru/ Basically, Google has an abysmal record on AI safety and ethics research, and is one of the big reasons they have struggled hiring researchers in LLMs, which in turn is the reason that despite massive capabilities they felt behind on the LLM tech to the point of starting to loose market in their core sector - search. Citing Google's PR on ethics and safety without providing context might indeed lead to a feeling of hope, but it is, unfortunately, a misplaced one...

Summary Following the news of secret documents leaks on Discord earlier this week, Bellingcat and to WaPo journalists decided to dig into the sources into the leaks and it looks like they gravitated around two minor youtube celebrities whose followers had a particular penchent for edgy shitposting. Based on the reports, the source is a community centered around a small youtuber (Oxide ~130k subs). historically famous for playing wargames and currently in the military. Essentially a community of guys into weapons, geopolitics and shitposting, the toxicity of some of his fans got to Oxide and the resulting bans pushed the most toxic subset of his fans to a separate forum, where the secret documents were posted over a span of months. From there, some of the documents were re-posted to a discord server of a self-proclaimed "shitposting internet microcelebrity" WowMao, from which they were reposted to a general Minecraft server - Minecraft Earth Map - before making their way to 4chan and Russian Milblogger telegram channel, who promptly re-posted them after editing Ukrainian casualties numbers. The WaPo articles are painting the image of a community of lonely teens and young adults who congregated to discord servers during he pandemic to keep socializing and in the race to appear cool one to another decided to post secret documents they had access to through work (or declare they won't rat out their internet friends who did that, on-record, in clear-voice to a major news outlet). Quotes Belligcat from 9th of April: WaPo from 12th of April 1: WaPo from 12th of April 2: My thoughts Looks like kids who grew up on the Navy Seal Copypasta are finally adults and after failing to become the actual Navy Seals are trying to pass of as ones among their online friend by leaking classified documents. I wonder how long before discords one is part of and youtuber one follows are join War Thunder in the latest questions on security clearances list and where LinusTechTips the LTT Minecraft Network would rank in that list. I also wonder, what tips would be needed for the parent who gave a consent for their kid to be interviewed by WaPo on-record. Especially when their kid is saying they new the identity of the currently most wanted person in the USA, but won't give it to the authorities. That also gives a more credence to Linus' obsession with what his kids do on the internet and who they interact with. Wonder if he would have additional tips for other parents wrt that. Sources https://www.washingtonpost.com/national-security/2023/04/12/discord-leaked-documents/ https://www.washingtonpost.com/technology/2023/04/12/discord-documents-oxide-wow-mao/ https://www.bellingcat.com/news/2023/04/09/from-discord-to-4chan-the-improbable-journey-of-a-us-defence-leak/ https://cybervillains.com/@alex/110189659142649808

Summary Cyber-Defense campus, a research arm of the Swiss DoD in everything related to cyber-security and cyber-defense, has released a first version of its report on the impact Large Language Models could have on cyber-security. After an introduction of the tech behind LLM, reviewing current models, and citing some fundamental limitations, they list the main threats they see, with the ones most relevant to the LTT audience being: - Phishing, especially when they have already compromised a system and can use Microsoft Office 365 Copilot - Ability to search better and deeper and have summarization of findings - Vulnerability of the code generated by them, unlikely to be detected by beginner coders most likely to use them to write code - Private information leakage from interactions with models when they are learning from interactions - Hijacking systems that are controlled by LLMs (a bit like SQL injections) In the part where they review Bing Chat and GPT-4, they cite and link the Feb 10th WAN show with a timestamp as the source for their evaluation of abilities and the most likely architecture and training modes used to achieve it. Quotes My thoughts Looks like a report that falls straight into the intersection of the currently two hot topics for LTT over the last couple of months. I would have been curious to hear Luke's take on the subject. Also, watching professionally and with straight face Linus and Luke going nuts live for almost an hour, Two Soyjacks style, might have been a new experience for some of them. Sources https://arxiv.org/abs/2303.12132

Summary While investing into OpenAI to get an edge in the LLM-based chatbots space, Microsoft silently lays off its AI safety and ethics team. Quotes My thoughts Looks like Bing Chat gaslighting and telling Luke it loves him is not a bug but a feature and going forwards, more and more AI products, especially from Microsoft, will be behaving in "interesting" ways. Sources https://techcrunch.com/2023/03/13/microsoft-lays-off-an-ethical-ai-team-as-it-doubles-down-on-openai/ https://www.independent.co.uk/tech/microsoft-layoff-ai-ethics-team-b2300237.html

I don't think you (as in LTT writers team and most LLT viewers) realize just how bad it is. At 13.5B parameters, it can fit into a single RTX4090/4080/3090Ti in float8/int8 mode, and 4chan already got a whiff of just how much all-but-impossible-to-detect gaslighting LLMs can bring along when some asshat ran no-approval no-consent LLM interaction experiments on them We are about to witness new depth of toxicity, harassment, catphishing and just phishing basically everywhere...

Understandable, but sets a shitty precedent though, especially for indie games in open access. And damage from YT strikes to smaller channels is already there, unfortunately.

I think what's missing in it is that there is a lot of collateral damage to the small YT creators who poured thousands of hours into understanding the game and making it fun for the their viewers. Like for instance the one that is mentioned in Raptor's video @00FF00 posted on Sunday. The fact that he can permanently loose his channel and his community of 10 000 viewers over an abusive DMCA strike and YT reliance on bots to remove offending content no questions asked is a massive problem for channels that do not have dedicated representatives is just plain stupid and something YouTube needs to change if they want to keep their position of preferred plateform of creators.

That's a bit extreme and will be disrupted by cats napping on top of it. That's more than a bit extreme, and also I don't want to get my eyes clawed out the next night. I wish, but HVAC is loud and annoying here and in winter drops humidity below 20%. Sigh. Already do it, but it does not seem to help in any meaningful way. Was figuring out I would have to do it, but had a faint hope I could find a case to save me some time and effort.

After seeing the pristine state of Jake's PC interiors upgrade during the Extreme Intel Upgrade, I went ahead and checked mine (CoolerMaster Silencio 452). And it wasn't pretty. Despite being only a couple of months in the apartment, and despite me vacuuming my fluffy cats' fluff around the house pretty much non-stop, some started making it into the case. What is Jake using to stop it in his? Or what alternative cases (hopefully low-noise) are there that are good at keeping the dust and animal fluff out of the PC?

Surprisingly, nope. Even if we go around the whole topic of writing and running code on custom hardware, the modern models are gigantic and require a lot of vRAM in inference alone; training is even worse than that. The current recommendation for ML are RTX4090s in local inference and fast loop development; H100s for on-premises and Azure otherwise. Besides, for price-performance ratio, GPUs have been all the rage since at least 2019. In 2020 Nvidia claimed to have trained at 540B parameters model, whereas Google had to write a new scheduler and dataflow layer to do the same on their TPUs in 2022.