Tech Time by Tim #49

Tech Time by Tim #49

Ever heard of a WUI? Now you have, even nature apparently has a User Interface now! It’s in need of an overhaul thanks to climate change, so we’d better make up our minds quickly. On that road to improvement, we’ll spot some car server crashes and printer rebellions but may also learn to feel the beat of our travel music in a more literal sense. Last but not least, a look at an LLM X-factor, and some Linus ‘Tact’ Tips.


At A Glance

A quick overview of this week’s content

·?????? The Week that Was: New WUI Design Principles, Make Up Your Mind

·?????? Go, Go Gadget: Car Server Crashes And Printer Rebellions, Feel The Beat

·?????? Rules of Engagement: The LLM X-Factor, Linus Tact Tips


The Week that Was:

A look back at the tech world of the past week

New WUI Design Principles:

There was something weird about the major Lahaina wildfires of Hawaii. Modern cities aren’t supposed to burn like that anymore, owing to advances in urban planning, building materials, and firefighting technologies. In reading up on why this particular Hawaiian city did burn ‘like that,’ I came upon the term ‘wildland-urban interface.’

The wildland-urban interface (WUI) isn’t just the border between nature and human settlements, it is the means through which these two opposing spaces interact with one another. WUI design then, could be considered as one of the main reasons why modern cities, up until now, have not had nearly the problems with fire that olden-day human settlements did. Now I’m not saying that Emperor Nero wouldn’t have benefitted from some fire safety classes, but I am saying that recent legal wins by climate activists could be beneficial for updating our WUI design principles.

Countries around the world are struggling with record temperatures, leading to problems such as excess wildfire smoke inhalation and, somewhat ironically, ‘hot’ sauce shortages. People have been testifying about climate change since at least the 1980’s yet here we are. Our computing demands keep getting heavier, our planet keeps getting hotter, and those datacentres aren’t going to cool themselves. There’s some hope on the horizon that computing might be able to clean up its act, which would at least go some way to lessening the water shortages that make bad fires a lot worse. LINK

Make Up Your Mind:

Three major things happened with regard to social media algorithms this past week. TikTok has started efforts to let European users turn off its algorithm, the Associated Press has published its guidelines for AI use by its journalists, and the Biden Administration has moved to block Texas’ Controversial Media law.

TikTok owes its rapid rise to its highly addictive algorithm. It is being forced to make that algorithm opt-in due to pending EU legislation. It’s doubtful very many people will, because if you could simply opt out of addiction, no one would ever need rehab. The gesture though serves as quite an effective move against the state of Texas, whose sweeping social media legislation has now drawn the attention of the federal government. Meanwhile, the AP, as one of the most reputable news outlets in the world, stands to exert a strong influence on the broader stance of respectable journalistic outlets towards usage of AI in their workflows.

‘Mainstreamlining’ is a word I use from time to time, and it is relevant here as well. What it entails is that you have complex technology or concepts that have existed for a long time on one side. On the other side is a need to make these things understandable for a wider public. Large, mainstream organisations such as the NYT are directly financially dependent on making complex subject matter understandable for a wider public. And as we’ve seen, a successful formula is swiftly replicated, hence the term ‘mainstreamlining.’ When sufficiently understood, the conversation can move into new territory, for example, talk about ‘heritage’ algorithms, relating specifically to art. LINK


Go, Go Gadget:

Experimental technologies and newly released gadgetry that are worth keeping an eye on

Car Server Crashes And Printer Rebellions:

Autopilot technology continues to be extensively tested in China with some interesting results thus far. At the same time, car theft continues to get more technologically advanced as well. Many of today’s ‘always online’ cars these days may not need cybercrime to take users for a ride though, at least if recent issues with 3D printing company Bambu’s cloud services are any indication. Let’s dive into that!

The race for autopilot dominance in China is important because the size and complexity of large, crowded Chinese cities is an excellent baseline to see just how reliable the technology is in general. It is also particularly important to car companies because China is one of their most important markets. The more versatile and expensive these vehicles become, the more lucrative they become for criminals.

In that context, the curious case of the rogue 3D printers may not seem relevant at first glance but it’s actually a great case study of overreliance on complex technical infrastructure. I’m being sarcastic when I say there’s absolutely no historical precedent of high-tech cars either locking users out or in their cars due to weird errors on the manufacturer’s end. Rebelling against their owners because something went screwy on the manufacturer’s servers.

Now, imagine if that car wasn’t a car at all, but a tank, what if your smart tank locked you out on a whim and decided to go on an adventure? That’d make for a pretty awkward conversation with command. It also points to a less often discussed aspect of the semi-autonomous and fully autonomous military vehicle debate. A lot these machines will at least initially be similar to the vehicles we already know and use today, and if we can’t get self-driving or printing right yet on consumer vehicles. LINK

Feel The Beat:

Some people really, really like listening to white noise, it’s a kind of ASMR for them. ASMR stands for Autonomous Sensory Meridian Response. It is a very particular sound, played on a loop for hours at a time. It’s so popular because ASMR is a kind of tingling sensation which starts at the scalp and then moves down the back of the neck and upper spine. Which is to say that this is basically that shivers down your spine sensation, but on demand. People into this style of content are really into it for that very reason, it’s almost like a drug for them, making it a massively lucrative niche as far as content creation goes. Supposedly ASMR creators can make $18000 dollars per month on Spotify. Money Spotify would rather not spend on the acoustic equivalent of hot air as far as investors are concerned.

Indeed, whilst ASMR continues to be relatively niche, yet the amount of revenue it is pulling away from more mainstream content is a thorn in the side of streaming services such as Spotify. Having said that, the neurological principles behind this phenomenon might be able to go mainstream via haptic suits such as this one. The goal of such suits is for any wearer to ‘feel the beat’ of their favourite songs. Honestly, that seems like it’d be a pretty good value proposition for audiophiles or for those seeking to have a festival-like experience at home.

Neuroscientists even recently succeeded in recreating a Pink Floyd song from the brain activity of a test subject. Imagine coupling a haptic suit to brainwaves in a similar manner to fine-tune the experience. To make that commercially viable there are a lot of practical and ethical hurdles to clear though. For example, with regard to ‘mental privacy.’ ASMR, if looked at uncharitably, can already be viewed as a kind of brain hack, so any future tech based on it would need to have reliable frameworks to protect the privacy and safety of what I’ll call ASMR 2.0 enjoyers. LINK ?


Rules of Engagement:

Ethics and legal matters regarding tech engagement

The LLM X-Factor:

Netflix has done a small game streaming field test, whilst X (formerly Twitter) got caught sabotaging links (again) to various different publications. Meanwhile, the New York Times (NYT) is considering suing OpenAI as the NYT makes major terms of service changes in order to explicitly forbid NYT content from being used to train Large Language Models (LLM’s).

Netflix testing a game streaming service is significant because it not only shows the cultural footprint of video games becoming too big for video streaming companies to ignore but also the cracks in the video streaming world’s piggybank widening as the Hollywood strikes carry on. Speaking of cracks in the piggybank, that’s pretty much why X keeps trying to prevent users from posting links. It doesn’t want people to leave the app, because that’s money X isn’t making from those users.

Both Netflix and X are concerned about user retention, hence why they want to keep as many users within their walls as they can. The NYT situation, whilst similar, also has the added wrinkle of brand protection. You see, back in April The Washington Post published research that seemed to indicate that the NYT was actually the 4th largest source of training data for popular LLM’s such as GPT.

Major news outlets such as the aforementioned Washington Post and Associated Press have already or are in the process of, making special licensing deals with generative AI companies. The NYT is notably not doing so, indicating that it likely feels confident enough in its chances for an exclusive deal or…perhaps it has something of its own cooking in the algorithmic kitchen. LINK

Linus Tact Tips:

Two big tech shakeups as far as popular communities go. Match Group (MG) has lost the non-profit that did background checks for its dating app Tinder. Popular YouTube Channel Linus Tech Tips (LTT) also got caught up in a journalistic integrity controversy. Both situations could have easily been avoided if either MG or LTT founder Linus Sebastian had shown more ‘tact’ and oh would you look at that? Guess we’ve got our heading for this item!

Both controversies are, essentially, breaches of social contract. Such a contract entails a series of written and unwritten agreements between groups of people that they behave themselves a certain way, refrain from behaving in other ways, and in return can reap certain benefits. For dating apps like Tinder, part of this contract is that you should be able to trust that your date won’t be a predator. In the case of a channel like LTT the contract focuses on getting trustworthy information that’s free of bias and hidden agendas.

Fans trust their favourite brands and influencers to be honest with them, rewarding them with attention and money. When it turns out that they were lied to, tempers flare due to the sense of betrayal many ‘fans’ feel. Which is what makes these particular cases interesting. For MG it was their partner, for LTT it was the community that felt taken for granted. The Non-profit was fine with doing the work for MG, but not if it’s mission to protect people from predators wasn’t taken seriously.

As for LTT, fans and rivals alike were fine with the channel’s dominant position in the tech space. To a certain extent at least. What they were not fine with was how much alleged amateurism and sloppiness ran rampant at LTT, despite how large and influential the organisation behind it had long since become. MG I honestly have little faith in, but depending on how LTT resolves it’s present predicament, it could be a major win for tech safety. It could set a positive example with regards to accountability and transparency. LINK


A Nice Cup of Serendipity:

Cool bits and bobs from around the web

Chinese Ghost Town LINK

Lunar Living LINK

Growing Teeth LINK

Hear Ye, Hear Ye LINK

Click Work LINK

Synthetic Data LINK

Collab Burnout LINK

Dave’s Dividends LINK

Finding Funding LINK


The Deep End:

A weekly batch of longform content recommendations

Bot Brains:

What might ChatGPT’s hypothetical brain look like? LINK

Open Source Musing:

An MIT tech review item on the complexities of Open-Source in our current climate of consolidation and proprietary tech. LINK

Deep Sea Beebe:

William Bebe’s account of becoming the first human being to see the deep ocean. LINK

iMac’s 25 Year Anniversary:

Not one, but a collection of articles commemorating the history of the iMac. Pick your poi…Apple.

·?????? Saved Apple

·?????? Visual History

·?????? Student Experience

·?????? Looking Ahead


One More Thing…

Guess who just got a new 3D printer from Bambu? My company! Well, we ordered it, along with another brand of 3D printer. They’ve still got to be delivered though. Boy, I sure hope I don’t wake up to the surviving members of our Tech Ops department fighting for their lives against a legion of ‘Printerminators.’ I’ve got the mental image of dreadful machines marching through a hail of nerf blaster fire as HR desperately looks for a colleague named Sarah to save us all. So yeah, the 3D printer issue wasn’t that big a problem in the grand scheme of things. The problems were fixed and an embarrassed Bambu apologised. Yet the fact that it was called a ‘rebellion’ by some cheeky reporters did spark my imagination because of the Terminator connection.

Imagine, waking up to your expensive 3D printer all out of resin, or whatever material it happens to use, and a whole bunch of things just standing there…menacingly! Connected devices always have to get a name assigned to them for easy identification, a lot of the time, you can manually change these to suit your needs or preferences. As such, as soon as we receive the 3D printers we’ve ordered, I’m putting forward the motion to call at least one of them ‘The Printerminator.’

I’m sure nothing could possibly go wrong with that; not like I’ll call the local Wi-Fi network ‘Skynet’ or anything. Having said that, let’s hope the company wide push towards more ‘destructive innovation’ won’t end up becoming too literal in the unlikely case that something does go slightly cyber pear shaped. If we have a repeat of the ‘Great Nerf War’, Finance will definitely want to have words again…

Tim Groot, Tech Time by Tim author.


要查看或添加评论,请登录

Triple | Part of Hypersolid的更多文章

社区洞察

其他会员也浏览了