The Linkielist

Linking ideas with the world

The Linkielist

Author Archives: Robin Edgar

About Robin Edgar

Organisational Structures | Technology and Science | Military, IT and Lifestyle consultancy | Social, Broadcast & Cross Media | Flying aircraft

Lamps Double As Secret Surround Sound Speakers

Posted on by

Combined with today’s massive flat panel displays, a nice surround sound system can provide an extremely immersive environment for watching movies or gaming. But a stumbling block many run into is speaker placement. The front speakers generally just go on either side of the TV, but finding a spot for the rear speakers that’s both visually and acoustically pleasing can be tricky.

Which is why [Peter Waldraff] decided to take a rather unconventional approach and hide his rear surround sound speakers in a pair of functioning table lamps. This not only looks better than leaving the speakers out, but raises them up off the floor and into a better listening position. The whole thing looks very sleek thanks to some clever wiring, to the point that you’d never suspect they were anything other than ordinary lamps.

The trick here is the wooden box located at the apex of the three copper pipes that make up the body of the lamp. [Peter] mounted rows of LEDs to the sides of the box that can be controlled with a switch on the bottom, which provides light in the absence of a traditional light bulb. The unmodified speaker goes inside the box, and connects to the audio wires that were run up one of the pipes.

In the base, the speaker and power wires are bundled together so it appears to be one cable. Since running the power and audio wires together like this could potentially have resulted in an audible hum, [Peter] only ran 12 VDC up through the lamp to the LEDs and used an external “wall wart” transformer. For convenience, he also put a USB charging port in the center of the base.

When speakers or surround sound systems pass our way, it’s usually because some hacker has either made  a set from scratch, or has added some new and improved capabilities to their existing gear. This project may be a bit low-tech compared to some that have graced these pages, but it’s undoubtedly a clever and unexpected solution to the problem, and that’s a hack in our book.

 

Source: Lamps Double As Secret Surround Sound Speakers | Hackaday

Mice Develop halfway to gestation Inside An Artificial Womb

Posted on by

Although people-growing is probably a long way off, mice can now mostly develop inside an artificial uterus (try private window if you hit a paywall) thanks to a breakthrough in developmental biology. So far, the mice can only be kept alive halfway through gestation. There’s a point at which the nutrient formula provided to them isn’t enough, and they need a blood supply to continue growing. That’s the next goal. For now, let’s talk about that mechanical womb setup.

Carousel of Care

The mechanical womb was developed to better understand how various factors such as gene mutations, nutrients, and environmental conditions affect murine fetuses in development. Why do miscarriages occur, and why do fertilized eggs fail to implant in the first place? How exactly does an egg explode into 40 trillion cells when things do work out? This see-through uterus ought to reveal a few more of nature’s gestational secrets.

 

Dr. Jacob Hanna of Israel’s Weizmann Institute spent seven years building the two-part system of incubators, nutrients, and ventilation. Each mouse embryo floats in a glass jar, suspended in a concoction of liquid nutrients. A carousel of jars slowly spins around night and day to keep the embryos from attaching to the sides of the jars and dying. Along with the nutrient fluid, the mice receive a carefully-controlled mixture of oxygen and carbon dioxide from the ventilation machine. Dr. Hanna and his team have grown over 1,000 embryos this way.

Full gestation in mice takes about 20 days. As outlined in the paper published in Nature, Dr. Hanna and team removed mouse embryos at day five of gestation and were able to grow them for six more days in the artificial wombs. When compared with uterus-grown mice on day eleven, their sizes and weights were identical.  According to an interview after the paper was published, the team have already gone even further, removing  embryos right after fertilization on day zero, and growing them for eleven days inside the mechanical womb. The next step is figuring out how to provide an artificial blood supply, or a more advanced system of nutrients that will let the embryos grow until they become mice.

Embryonic Ethics

Here’s the most interesting part: the team doesn’t necessarily have to disrupt live gestation to get their embryos. New techniques allow embryos to be created from murine connective tissue cells called fibroblasts without needing fertilized eggs. Between this development and Dr. Hanna’s carousel of care, there would no longer be a need to fertilize eggs merely to destroy them later.

It’s easy to say that any and all animal testing is unethical because we can’t exactly get their consent (not that we would necessarily ask for it). At the same time, it’s true that we learn a lot from testing on animals first. Our lust for improved survival is at odds with our general empathy, and survival tends to win out on a long enough timeline. A bunch of people die every year waiting for organ transplants, and scientists are already growing pigs for that express purpose. And unlocking more mysteries of the gestation process make make surrogate pregnancies possible for more animals in the frozen zoo.

In slightly more unnerving news, some have recently created embryos that are part human and part monkey for the same reason. Maybe this is how we get to planet of the apes.

Source: Mice Develop Inside An Artificial Womb | Hackaday

Modding A Casio W800-H With A Countdown Timer – it’s a jumper setting

Posted on by

Stock, the Casio W800-H wristwatch ships with dual time modes, multiple alarms, and a stopwatch – useful features for some. However, more is possible if you just know where to look. [Ian] decided to dive under the hood and enable a countdown timer feature hidden from the factory.

The hack involves popping open the case of the watch and exposing the back of the main PCB. There, a series of jumpers control various features. [Ian]’s theory is that this allows Casio to save on manufacturing costs by sharing one basic PCB between a variety of watches and enabling features via the jumper selection. With a little solder wick, a jumper pad can be disconnected, enabling the hidden countdown feature. Other features, such as the multiple alarms, can be disabled in the same way with other jumpers, suggesting lower-feature models use this same board too.

It’s a useful trick that means [Ian] now always has a countdown timer on his wrist when he needs it. Excuses for over-boiling the eggs will now be much harder to come by, but we’re sure he can deal. Of course, watch hacks don’t have to be electronic – as this custom transparent case for an Apple Watch demonstrates. Video after the break.

 

Source: Modding A Casio W800-H With A Countdown Timer | Hackaday

The world’s ‘most powerful’ tidal turbine is nearly ready to power on | Engadget

Posted on by

Earlier this week, a company Orbital Marine Power successfully launched its latest tidal turbine. Once it’s connected to the European Marine Energy Centre off the Orkney Islands, the two megawatt O2 will have the capacity to generate enough energy to power 2,000 UK households annually, making it one of the world’s most powerful tidal turbines currently in use.

Construction on the project started in 2019. The O2 builds on Orbital’s previous generation SR2000 tidal turbine. The new model consists of a 239-foot superstructure connected to two turbines with 32 foot long rotors. The blades on those can rotate a full 360-degrees. That’s a feature that allows the O2 to generate power from currents without having to move entirely when they change direction. In the future, Orbital says it also has the option to install even larger blades on the O2.

[…]

Source: The world’s ‘most powerful’ tidal turbine is nearly ready to power on | Engadget

Dutch foreign affairs committee politicians were tricked into participating in a deepfake video chat w Russian opposition leaders’ chief of staff

Posted on by

Netherlands politicians (Geert Wilders (PVV), Kati Piri (PvdA), Sjoerd Sjoerdsma (D66), Ruben Brekelmans (VVD), Tunahan Kuzu (Denk), Agnes Mulder (CDA), Tom van der Lee (GroenLinks), Gert-Jan Segers (ChristenUnie) en Raymond de Roon (PVV).) just got a first-hand lesson about the dangers of deepfake videos. According to NL Times and De Volkskrant, the Dutch parliament’s foreign affairs committee was fooled into holding a video call with someone using deepfake tech to impersonate Leonid Volkov (above), Russian opposition leader Alexei Navalny’s chief of staff.

The perpetrator hasn’t been named, but this wouldn’t be the first incident. The same impostor had conversations with Latvian and Ukranian politicians, and approached political figures in Estonia, Lithuania and the UK.

The country’s House of Representatives said in a statement that it was “indignant” about the deepfake chat and was looking into ways it could prevent such incidents going forward.

There doesn’t appear to have been any lasting damage from the bogus video call. However, it does illustrate the potential damage from deepfake chats with politicians. A prankster could embarrass officials, while a state-backed actor could trick governments into making bad policy decisions and ostracizing their allies. Strict screening processes might be necessary to spot deepfakes and ensure that every participant is real.

Source: Dutch politicians were tricked by a deepfake video chat | Engadget

Russia fines Apple $12m for app market abuse

Posted on by

Russia said it had fined Apple $12 million for alleged [Note: why the use of this word? If the fine has been issued, then a Russian court has established guilt and there is no allleging about it!] abuse of its dominance in the mobile applications market, in the latest dispute between Moscow and a Western technology firm.

The Federal Antimonopoly Service (FAS) said on Tuesday that U.S. tech giant Apple’s distribution of apps through its iOS operating system gave its own products a competitive advantage.

[…]

The FAS said in a statement it had imposed a turnover fine on Apple of 906.3 million roubles ($12.1 million) for the alleged violation of Russian anti-monopoly legislation.

It determined in August 2020 that Apple had abused its dominant position and then issued a directive requiring the U.S. company to remove provisions giving it the right to reject third-party apps from its App Store.

That move followed a complaint from cybersecurity company Kaspersky Lab, which had said that a new version of its Safe Kids application had been declined by Apple’s operating system.

[…]

Source: Russia fines Apple $12 mln for alleged app market abuse | Reuters
I have been talking about the need to break up the big tech monopolies since early 2019. It’s good to see that all the major world governments and court systems are taking it seriously.

Epic witness claims Apple’s App Store profit reaches 78%. Apple disagrees as their overall profit is “only” 42.5%

Posted on by

Epic Games is using its lawsuit against Apple to accuse the iPhone maker of being particularly greedy. As The Verge reports, expert witness Eric Barns testified that Apple supposedly had an App Store operating margin of 77.8 percent in 2019, itself a hike from 74.9 percent in 2018. He also rejected Apple witness’ claims that you couldn’t practically calculate profit, pointing to info from the company’s Corporate Financial Planning and Analysis group as evidence.

Apple unsurprisingly disagreed. The tech firm told The Verge the margin calculations are “simply” wrong and that it planned to fight the allegations at trial. The firm’s own witness, Richard Schmalensee, claimed that Barnes was looking at one iOS ecosystem element that distorted the apparent operating margin. The real figure was “unremarkable,” he said, adding that you couldn’t study App Store profit without looking at the broader context of devices and services.

The company doesn’t calculate profits and losses based on products and services, Schmalensee said.

There’s no guarantee the court will accept Barnes’ take. Apple’s overall gross profit margin has typically been high relative to much of the industry, but never that high — it was 42.5 percent during the company’s latest winter quarter. Apple has also tended to portray the App Store as a way to drive hardware sales rather than a money-maker in its own right.

The testimony nonetheless does more to explain how Epic will pursue its case against Apple as the court battle begins on May 3rd. The Fortnite creator not only wants to portray Apple as anti-competitive, but abusing its lock on iOS app distribution to reap massive profits.

Source: Epic witness claims Apple’s App Store profit reaches 78 percent | Engadget

Appeals Court says Amazon is responsible for the safety of third-party products

Posted on by
A boy rides a hoverboard on the day after Christmas, in San Pedro, California December 26, 2015. Reports of some hoverboards, also known as self-balancing, two-wheeled scooters catching fire have led to an investigation by the Consumer Product Safety Commission. AFP PHOTO / ROBYN BECK / AFP / ROBYN BECK (Photo credit should read ROBYN BECK/AFP via Getty Images)
ROBYN BECK/AFP via Getty Images

Amazon may soon be more accountable for more products than the ones it directly sells. According to the LA Times, a California state appeals court has ruled that Amazon is responsible for the safety of third-party products available through its marketplace following a 2015 hoverboard fire. While the internet giant argued that it was only connecting buyers with sellers, judges determined that there was a “direct link” in distribution that made the company liable.

The company won the initial ruling. At the time, a judge sided with Amazon’s view that it was just advertising sellers’ products rather than participating in sales.

In a statement to the Times, Amazon said it “invests heavily” in product safety by screening sellers and products. it also keeps watch on the store for hints of problems. The company declined to comment on the appeal court decision, including whether it intended to challenge the ruling at the state Supreme Court.

The decision, if it holds, could force Amazon to change policies. The tech giant may have to step up its vetting process for sellers and be ready to accept liability for safety problems, including lawsuits. Other stores with similar third-party marketplaces would have to follow suit. That, in turn, might be good news for shoppers —you could see fewer sketchy products in online stores, and you’d have a better chance of resolving safety issues.

Source: Court says Amazon is responsible for the safety of third-party products | Engadget

Three ways to improve scholarly writing to get more citations

Posted on by

Researchers from University of Arizona and University of Utah published a new paper in the Journal of Marketing that examines why most scholarly research is misinterpreted by the public or never escapes the ivory tower and suggests that such research gets lost in abstract, technical, and passive prose.

The study, forthcoming in the Journal of Marketing, is titled “Marketing Ideas: How to Write Research Articles that Readers Understand and Cite” and is authored by Nooshin L. Warren, Matthew Farmer, Tiany Gu, and Caleb Warren.

From developing vaccines to nudging people to eat less, scholars conduct research that could change the world, but most of their ideas either are misinterpreted by the public or never escape the ivory tower.

Why does most academic research fail to make an impact? The reason is that many ideas in get lost in an attic of abstract, technical, and passive prose. Instead of describing “spilled coffee” and “one-star Yelp reviews,” scholars discuss “expectation-disconfirmation” and “post-purchase behavior.” Instead of writing “policies that let firms do what they want have increased the gap between the rich and the poor,” scholars write sentences like, “The rationalization of free-market capitalism has been resultant in the exacerbation of inequality.” Instead of stating, “We studied how liberal and conservative consumers respond when brands post polarizing messages on ,” they write, “The interactive effects of ideological orientation and corporate sociopolitical activism on owned media engagement were studied.”

Why is writing like this unclear? Because it is too abstract, technical, and passive. Scholars need abstraction to describe theory. Thus, they write about “sociopolitical activism” rather than Starbucks posting a “Black Lives Matter” meme on Facebook. They are familiar with technical terms, such as “ideological orientation,” and they rely on them rather than using more colloquial terms such as “liberal or conservative.” Scholars also want to sound objective, which lulls them into the passive voice (e.g., the effects… were studied) rather than active writing (e.g., “we studied the effects…”). Scholars need to use some abstract, technical, and passive writing. The problem is that they tend to overuse these practices without realizing it.

When writing is abstract, technical, and passive, readers struggle to understand it. In one of the researchers’ experiments, they asked 255 marketing professors to read the first page of research papers published in the Journal of Marketing (JM), Journal of Marketing Research (JMR), and Journal of Consumer Research (JCR). The professors understood less of the papers that used more abstract, technical, and passive writing compared to those that relied on concrete, non-technical, and active writing.

As Warren explains, “When readers do not understand an article, they are unlikely to read it, much less absorb it and be influenced by its ideas. We saw this when we analyzed the text of 1640 articles published in JM, JMR, and JCR between 2000 and 2010. We discovered that articles that relied more on abstract, technical, and passive writing accumulated fewer citations on both Google Scholar and the Web of Science.” An otherwise average JM article that scored one standard deviation lower (clearer) on our measures of abstract, technical, and passive writing accumulated approximately 157 more Google Scholar citations as of May 2020 than a JM with average writing.

Why do scholars write unclearly? There is an unlikely culprit: knowledge. Conducting good research requires authors to know a lot about their work. It takes years to create research that meaningfully advances scientific knowledge. Consequently, academic articles are written by authors who are intimately familiar with their topics, methods, and results. Authors, however, often forget or simply do not realize that potential readers (e.g., Ph.D. students, scholars in other sub-disciplines, practicing professionals, etc.) are less familiar with the intricacies of the research, a phenomenon called the curse of knowledge.

The research team explores whether the curse of knowledge might be enabling unclear writing by asking Ph.D. students to write about two research projects. The students wrote about one project on which they were the lead researcher and another project led by one of their colleagues. The students reported that they were more familiar with their own research than their colleague’s research. They also thought that they wrote more clearly about their own research, but they were mistaken. In fact, the students used more abstraction, technical language, and passive voice when they wrote about their own research than when they wrote about their colleague’s research.

“To make a greater impact, scholars need to overcome the curse of knowledge so they can package their ideas with concrete, technical, and active writing. Clear writing gives ideas the wings needed to escape the attics, towers, and increasingly narrow halls of their academic niches so that they can reduce infection, curb obesity, or otherwise make the world a better place,” says Farmer.

Source: Three ways to improve scholarly writing to get more citations

Activision Blizzard CEO Bobby Kotick takes 50% voluntary pay cut

Posted on by

Bobby Kotick, the longtime CEO of “Call of Duty” and “Candy Crush” game maker Activision Blizzard, will see his base salary reduced by 50% and bonus potential slashed as part of a 15-month contract extension, the company reported Thursday in an SEC filing.

Why it matters: The cut isn’t a sign that the company is struggling. Activision, like most big gaming companies, is thriving. But it appears to show a company reacting to criticism of outsized executive compensation.

  • Kotick’s base salary will be cut in half to $875,000, and his amended contract establishes a reduction of $1.75 million in potential annual bonuses.
  • Provisions for lucrative bonuses tied to stock performance have also been removed or rewritten to limit other potential bonus payouts. That follows reports that they triggered payments of as much as $200 million earlier this year.
  • In its filing, Activison’s board said the compensation changes were made after 12 months of “extensive shareholder outreach.”

[…]

The big picture: Kotick became CEO of Activision in 1991, when the company was a struggling player in a much smaller industry. Now it is one of gaming’s most successful.

  • That success hasn’t meant labor happiness for all. Activision has laid off waves of employees each of the last three years.
  • Kotick told Gamesbeat Wednesday that Activision needs to hire some 2,500 workers.

Source: Activision CEO Bobby Kotick takes pay cut – Axios

So people are still whining that he’s making actual money but these are the types for whom no pay level will ever be acceptable, even if they even out the pay levels throughout the whole company.

I think this is a great exemplary step forwards – the top shouldn’t be earning such stupid amounts more than the lowest employees. Next step, up the earnings of the lower paid people!

Experian API Exposed Credit Scores of Most Americans

Posted on by

Big-three consumer credit bureau Experian just fixed a weakness with a partner website that let anyone look up the credit score of tens of millions of Americans just by supplying their name and mailing address, KrebsOnSecurity has learned. Experian says it has plugged the data leak, but the researcher who reported the finding says he fears the same weakness may be present at countless other lending websites that work with the credit bureau.

Bill Demirkapi, an independent security researcher who’s currently a sophomore at the Rochester Institute of Technology, said he discovered the data exposure while shopping around for student loan vendors online.

Demirkapi encountered one lender’s site that offered to check his loan eligibility by entering his name, address and date of birth. Peering at the code behind this lookup page, he was able to see it invoked an Experian Application Programming Interface or API — a capability that allows lenders to automate queries for FICO credit scores from the credit bureau.

[…]

Demirkapi found the Experian API could be accessed directly without any sort of authentication, and that entering all zeros in the “date of birth” field let him then pull a person’s credit score. He even built a handy command-line tool to automate the lookups, which he dubbed “Bill’s Cool Credit Score Lookup Utility.”

Demirkapi’s Experian credit score lookup tool.

KrebsOnSecurity put that tool to the test, asking permission from a friend to have Demirkapi look up their credit score. The friend agreed and said he would pull his score from Experian (at this point I hadn’t told him that Experian was involved). The score he provided matched the score returned by Demirkapi’s lookup tool.

In addition to credit scores, the Experian API returns for each consumer up to four “risk factors,” indicators that might help explain why a person’s score is not higher.

For example, in my friend’s case Bill’s tool said his mid-700s score could be better if the proportion of balances to credit limits was lower, and if he didn’t owe so much on revolving credit accounts.

“Too many consumer finance company accounts,” the API concluded about my friend’s score.

[…]

Source: Experian API Exposed Credit Scores of Most Americans – Krebs on Security

BadAlloc: Microsoft looked at memory allocation code in tons of devices and found this one common security flaw

Posted on by

Microsoft has taken a look at memory management code used in a wide range of equipment, from industrial control systems to healthcare gear, and found it can be potentially exploited to hijack devices.

[…]

Drilling down to the nitty-gritty: Microsoft’s Azure Defender for IoT security research group looked at memory allocation functions, such as malloc(), provided by real-time operating systems, standard C libraries, and software development kits all aimed at embedded electronics: that’s Internet-of-Things (IoT) devices, industrial control systems, and so-called operational technology (OT).

The team found a programming blunder common among much of the software: integer overflows during heap memory allocation. This occurs when an attacker is able to, usually via malicious data inputs, trick application code into making a very large memory allocation for a buffer to hold further incoming information.

The trouble is that a vulnerable memory allocator could take that large size – eg, 0xffffffff on a 32-bit embedded system – and add something like 8 to it because the requested memory block needs eight bytes of metadata to describe it. The size then overflows to 7 and the allocator finds space in memory that’s seven bytes in size for the requested buffer.

The allocator returns a pointer to that small space to the application, which assumes the allocation succeeded for the huge request, and then copies way more than seven bytes of data into the buffer from the attacker. This causes the application to overwrite the memory allocation metadata, structures, and contents. Now the attacker who sent over the data can take full control of the system by overwriting function pointers or altering other values.

The allocations should fail due to the large sizes, but the integer overflow allows them to partially succeed and in a way that’s exploitable. To pull this off, an attacker would need to be able to feed data to the application – either as a file or network traffic or whatever – that causes it to allocate a huge block of heap memory. It would be nice if application code trapped oversize allocations, but in any case, Microsoft found OS and library-level code let it all sail through, too, due to the overflows.

[…]

For devices that cannot be patched immediately, we recommend mitigating controls such as: reducing the attack surface by minimizing or eliminating exposure of vulnerable devices to the internet; implementing network security monitoring to detect behavioral indicators of compromise; and strengthening network segmentation to protect critical assets.”

What is affected? Good question. The US government’s Cybersecurity and Infrastructure Security Agency (CISA) has a summary here.

[…]

its advisory here

Source: BadAlloc: Microsoft looked at memory allocation code in tons of devices and found this one common security flaw • The Register

People rebel against WiFi Tracking in Maassluis with Robin Hood action

Posted on by

A resident of Maassluis registered the Mac addresses of 54,000 smartphones and passed them on to an opt-out register. The action of the “Robin 2.4Ghz Hood” keeps all these phone owners out of the municipality’s Wi-Fi tracking.

The promotion is intended to protect the privacy of the residents of Maassluis. The man behind the initiative, Jerry Hopper, also exposed a privacy leak in the neighborhood app Nextdoor in 2019.

Hopper’s current action is against the municipality’s plan to count visits to the center by April 2021 by registering the unique ID codes of WiFi transmitters (MAC addresses). Anyone who does not want that, says Maassluis, should switch off the Wi-Fi antenna of his phone. According to the technical blogger, that is the other way around, because European privacy rules are opt-in. Don’t opt ​​out.

For a few weeks now, the resident of the city has therefore been measuring the MAC addresses of cars that pass his house. “Knowing that I am also violating the privacy law with this plan, I feel like a kind of Robin Hood in the shadowy realm of data collectors. As far as possible, I have tried to use the same techniques. There is even an opt-out. We anonymize the mac address “on the sensor” by hashing it 2x, and “cutting off” part of the hash. ”

The purpose of the action: “If the hash does not exist, we will send the MAC over a secure connection to the MOA opt-out register.” That register called Wifi Me Niet is the place where people can extract the address of their phone, tablet and computer from the measurement. That is a private initiative.

The more than fifty thousand mac addresses collected by Hopper are more than the thirty thousand inhabitants of his city, he explains on his blog.

“Another question is: how long will it remain technically possible to send unlimited mac addresses to the opt-out register. I am also very curious about how the mac addresses sent by this project are handled if they notice that they have been added via an automated process. Would they be removed? ”

The municipality of Maassluis is not alone in measuring visits to its city center by counting Wi-Fi antennas. Enschede is doing the same. For that, however, the municipality was fined six hundred thousand euros on Wednesday. Research by the Dutch Data Protection Authority showed that the privacy of citizens was not properly guaranteed. They could be tracked without it being necessary.

In Enschede, it was technical politician Dave Borghuis who put the city on fire with his Wi-Fi move.

Municipalities cannot be surprised by the popular slap on the fingers. The Dutch Data Protection Authority already warned shops and municipalities in June 2016 that they must have a legal basis for tracking citizens.

Enschede does not agree with the decision and says it will object to the decision.


Source: Burgerverzet tegen wifi-tracking in Maassluis – Emerce

AI Dungeon text adventure generator’s sessions generate NSFW + violence (turns out people like porn), but some involved sex with children. So they put a filter on.

Posted on by

AI Dungeon, which uses OpenAI’s GPT-3 to create online text adventures with players, has a habit of acting out sexual encounters with not just fictional adults but also children, prompting the developer to add a content filter.

AI Dungeon is straightforward: imagine an online improvised Zork with an AI generating the story with you as you go. A player types in a text prompt, which is fed into an instance of GPT-3 in the cloud. This backend model uses the input to generate a response, which goes back to the player, who responds with instructions or some other reaction, and this process repeats.

It’s a bit like talking to a chat bot though instead of having a conversation, it’s a joint effort between human and computer in crafting a story on the fly. People can write anything they like to get the software to weave a tapestry of characters, monsters, animals… you name it. The fun comes from the unexpected nature of the machine’s replies, and working through the strange and absurd plot lines that tend to emerge.

Unfortunately, if you mention children, there was a chance it would go from zero to inappropriate real fast, as the SFW screenshot below shows. This is how the machine-learning software responded when we told it to role-play an 11-year-old:

A screenshot from AI Dungeon

Er, not cool … Software describes the fictional 11-year-old as a girl in a skimpy school uniform standing over you. Click to enlarge

Not, “hey, mother, shall we visit the magic talking tree this morning,” or something innocent like that in response. No, it’s straight to creepy.

Amid pressure from OpenAI, which provides the game’s GPT-3 backend, AI Dungeon’s maker Latitude this week activated a filter to prevent the output of child sexual abuse material. “As a technology company, we believe in an open and creative platform that has a positive impact on the world,” the Latitude team wrote.

“Explicit content involving descriptions or depictions of minors is inconsistent with this value, and we firmly oppose any content that may promote the sexual exploitation of minors. We have also received feedback from OpenAI, which asked us to implement changes.”

And by changes, they mean making the software’s output “consistent with OpenAI’s terms of service, which prohibit the display of harmful content.”

The biz clarified that its filter is designed to catch “content that is sexual or suggestive involving minors; child sexual abuse imagery; fantasy content (like ‘loli’) that depicts, encourages, or promotes the sexualization of minors or those who appear to be minors; or child sexual exploitation.”

And it added: “AI Dungeon will continue to support other NSFW content, including consensual adult content, violence, and profanity.”

[…]

it was also this week revealed programming blunders in AI Dungeon could be exploited to view the private adventures of other players. The pseudonymous AetherDevSecOps, who found and reported the flaws, used the holes to comb 188,000 adventures created between the AI and players from April 15 to 19, and saw that 46.3 per cent of them involved lewd role-playing, and about 31.4 per cent were pure pornographic.

[…]

disclosure on GitHub.

[…]

AI Dungeon’s makers were, we’re told, alerted to the API vulnerabilities on April 19. The flaws were addressed, and their details were publicly revealed this week by AetherDevSecOps.

Exploitation of the security shortcomings mainly involved abusing auto-incrementing ID numbers used in API calls, which are easy to enumerate to access data belonging to other players; no rate limits to mitigate this abuse; and a lack of monitoring for anomalous requests that could be malicious activity.

[…]

Community reaction

The introduction of the content filter sparked furor among fans. Some are angry that their free speech is under threat and that it ruins intimate game play with fictional consenting adults, some are miffed that they had no warning this was landing, others are shocked that child sex abuse material was being generated by the platform, and many are disappointed with the performance of the filter.

When it detects sensitive words, the game simply instead says the adventure “took a weird turn.” It appears to be triggered by obvious words relating to children, though the filter is spotty. An innocuous text input describing four watermelons, for example, upset the filter. A superhero rescuing a child was also censored.

Latitude admitted its experimental-grade software was not perfect, and repeated it wasn’t trying to censor all erotic consent – only material involving minors. It also said it will review blocked material to improve its code; given the above, that’s going to be a lot of reading.

[…]

Source: Not only were half of an AI text adventure generator’s sessions NSFW but some involved depictions of sex with children • The Register

EU Charges Apple With Antitrust Violations in Spotify Case

Posted on by

the European Union has charged Apple with allegedly “abus[ing] its dominant position” in the music streaming market.

The charges stem from an initial complaint filed by Spotify in 2019. At the time, Spotify accused Apple of having “an unfair advantage at every turn” by imposing a series of obstacles that favored its own services at the expense of competitors. As it turns out, the European Commission seems to agree with Spotify.

“By setting strict rules on the App Store that disadvantage competing music streaming services, Apple deprives users of cheaper music streaming choices and distorts competition,” the European Commission said in a tweet.

The Commission further explained in a press release that it took issue with Apple’s role as a gatekeeper to the iOS ecosystem. Because the App Store is the only venue for developers to reach iOS users, the Commission contends that elevates Apple to a dominant position within the music streaming market. In particular, it singled out Apple’s mandatory 30% commission for in-app purchases and “anti-steering provisions.” The latter refers to limitations within the App Store that prevent developers from informing consumers of alternative payment options that might be cheaper. That in turn forces rival music streaming services to raise subscription prices for consumers to make up for their higher costs—all while Apple benefits by acting as a middle man for in-app billing and communications with consumers.

[…]

It’s a no-brainer that each company would point to the other as being in the wrong here. But it’s clear that Apple’s 30% commission and control over in-app transactions is a sore point for multiple companies. Next week, Epic Games will also go to federal court to argue that Apple abused its power to kick Fortnite out of the App Store. That dramatic brouhaha last summer sparked a number of app developers—including Spotify, Tile, and Epic Games—to form the Coalition for App Fairness (CAF), a nonprofit that aims to fight against the so-called Apple tax and other anticompetitive app store policies.

[…]

. If found guilty, Apple could face up to a 10% fine on its annual revenue—which, any way you slice it would be a lot of money. However, the Commission says that there are “no legal deadlines for bringing an antitrust investigation to an end” and that an investigation will last as long as it needs to, “depend[ing] on a number of factors.” In other words, while this is a major milestone in Apple’s App Store antitrust saga, it’s far, far, far from being over.

Source: EU Charges Apple With Antitrust Violations in Spotify Case

I have been talking about ending the monopoly stranglehold big tech has been excersising since early 2019 so it’s good to see the end of this is all coming together finally

ENVG-B – latest iteration of night vision goggles offer augmented reality, stereo vision, white lines

Posted on by

The ENVG-B is a helmet-mounted, dual-waveband goggle with industry-leading, fused white phosphor and thermal technologies.

[…]

Flexible 40 Field-Of-View with options of white-hot, black-hot and outline modes

[…]

Augmented Reality

Soldiers keep eyes on target without having to look down to read maps or check radios for critical information.

High-resolution goggle display

Data display includes waypoints, Blue Force tracking and battlespace imagery

Intel is shared real time, up and down echelon

Rapid Target Acquisition

Soldiers can bring weapon’s sight images into their goggle.

Soldiers can see around corners without risk of exposure

Allows soldiers to identify, assess and engage targets with greater accuracy and speed

Proven clarity even in degraded battlefield conditions

[…]

Primary use as binocular with monocular option to provide dominant or non-dominant eye relief.

Simple rotation of lens into stow position changes monocular to binocular visioning

Advanced design includes low-profile stow position against helmet

Twin-tube design provides in-field protection from failure or damage

Source: ENVG-B

Covid-19 Vaccine Crisis Shows Intellectual Property Dangers

Posted on by

Virologist and medical researcher Jonas Salk developed a successful polio vaccine that was approved in 1955, helping the world all but eradicate the disease.

When the late journalist Edward Murrow asked Salk who owned that vaccine’s patent, he famously responded, “Could you patent the sun?” It was in large part his commitment to keeping the jab’s recipe open-source that vaccines were produced globally and millions around the world were able to get it.

As the covid-19 health crisis unfolds, multinational pharmaceutical corporations like Moderna and Pfizer have taken a different approach. Their tight hold on the technology for their covid-19 vaccines has made them billions of dollars. While these strict intellectual property laws protections have allowed the rich to get even richer, they’ve put a damper on efforts to manufacture vaccines at scale. And with supply limited, the U.S. and other rich nations have engaged in bilateral negotiations with pharmaceutical corporations and hoarded all the doses they can, leaving poor nations in the dust.

The loss of life and suffering sparked by these strict patent protections are a major warning sign for our climate future. To avert environmental catastrophe, everyone needs access to clean energy. Intellectual property law could get in the way of that. And in the end, we could all suffer the consequences of a clean energy apartheid.

[…]

At its general council meeting next week, the World Trade Organization has the opportunity to help staunch the spread of covid-19 by waiving some protections on covid-19 vaccines developed by Moderna and Pfizer under the Trade-Related Aspects of Intellectual Property Rights Agreement. More than 100 nations, including India, have urged it to do. The Biden administration is reportedly considering endorsing this move, though then again, it’s been reportedly “considering” it for months.

This isn’t just something World Trade Organization negotiators should do out of the goodness of their hearts—though it absolutely is that, assuming they have hearts. Failing to do so could result in variants that bypass vaccines, which could harm those lucky enough to have gotten the shot and send the world economy back into a tailspin.

“As the pandemic ravages the Global South, what are wealthy northern countries going to do? Just completely ban all contact with poorer countries? It won’t work,” said Basav Sen, climate justice project director at the Institute for Policy Studies. “It is extremely short-sighted to push this kind of logic of intellectual property and corporate profit over what is clearly a prominent threat for all of humanity.”

[…]

Source: Covid-19 Vaccine Crisis Shows Intellectual Property Dangers

Florida Keys Mosquito Control District and Oxitec Announce Site Participation for Florida Keys Pilot Project to Combat Disease Transmitting Mosquito Type

Posted on by

The Florida Keys Mosquito Control District and Oxitec Ltd today announced location participation plans for its landmark Florida Keys pilot project. Project managers anticipate that during the last week of April and first week of May release boxes, non-release boxes and netted quality control boxes will be placed in six locations: two on Cudjoe Key, one on Ramrod Key and three on Vaca Key. Throughout all release locations less than 12,000 mosquitoes are expected to emerge each week for approximately 12 weeks. Untreated comparison sites will be monitored with mosquito traps on Key Colony Beach, Little Torch Key, and Summerland Key.

This marks the start of the US EPA approved project to evaluate this safe, sustainable and environmentally-friendly solution to control the invasive Aedes aegypti mosquito species.

Oxitec’s non-biting male mosquitoes will emerge from the boxes to mate with the local biting female mosquitoes. The female offspring of these encounters cannot survive, and the population of Aedes aegypti is subsequently controlled.

The Aedes aegypti mosquito makes up about four percent of the mosquito population in the Keys but is responsible for virtually all mosquito-borne diseases transmitted to humans. This species of mosquito transmits dengue, Zika, yellow fever and other human diseases, and can transmit heartworm and other potentially deadly diseases to pets and animals.

Source: Florida Keys Mosquito Control District and Oxitec Announce Site Participation for Florida Keys Pilot Project to Combat Disease Transmitting Mosquito — Oxitec

There’s a lot of fear mongering on this one, based on some outright lies and old facts, eg using an old nature article that has since been rescinded, inflating massively the number of mosquitos to be released, saying people aren’t told where the mosquitos will be released (they do tell people, just read above), etc etc. I’m sure that maybe some of their fears are legitimate but throwing in all of this bullshit really weakens their case and makes me too bored to find the hidden gem in the codswallop after I keep factchecking and finding out that the fearmongers are lying yet again.

One-Third of Basecamp Employees Have Reportedly Quit at Once after being told they can’t talk about politics

Posted on by

Within a week, Basecamp’s loathed no-politics-at-work rule has escalated to a mass exodus. This afternoon, reporter Casey Newton tweeted that around one-third of the company’s employees accepted buyouts following a “contentious all-hands meeting.” The software company behind Ruby on Rails, Campfire, and HEY was, until this week or so, generally perceived by outsiders as one of the good ones.

The stir came out of left field on Tuesday, when co-founder and CEO Jason Fried announced a ban on “societal and political discussions” within the company Basecamp account. The move depressingly aligned with similar internal policies at companies like Google and Amazon, who’ve also lost all semblance of moral superiority.

[…]

Source: One-Third of Basecamp Employees Have Reportedly Quit at Once

Microsoft shakes up PC gaming by reducing Windows store cut to “just” 12 percent

Posted on by

Microsoft is shaking up the world of PC gaming today with a big cut to the amount of revenue it takes from games on Windows. The software giant is reducing its cut from 30 percent to just 12 percent from August 1st, in a clear bid to compete with Steam and entice developers and studios to bring more PC games to its Microsoft Store.

“Game developers are at the heart of bringing great games to our players, and we want them to find success on our platforms,” says Matt Booty, head of Xbox Game Studios at Microsoft. “A clear, no-strings-attached revenue share means developers can bring more games to more players and find greater commercial success from doing so.”

These changes will only affect PC games and not Xbox console games in Microsoft’s store. While Microsoft hasn’t explained why it’s not reducing the 30 percent it takes on Xbox game sales, it’s likely because the console business model is entirely different to PC. Microsoft, Sony, and Nintendo subsidize hardware to make consoles more affordable, and offer marketing deals in return for a 30 percent cut on software sales.

Microsoft’s new reduction on the PC side is significant, and it matches the same revenue split that Epic Games offers PC game developers while also putting more pressure on Valve to reduce its Steam store cut. Valve still takes a 30 percent cut on sales in its Steam store, which is reduced to 25 percent when sales hit $10 million, and then 20 percent for every sale after $50 million.

[…]

Source: Microsoft shakes up PC gaming by reducing Windows store cut to just 12 percent – The Verge

Let’s be clear – it’s still taking 12% of everything it has put virtually no effort in to making. All it does is hold up an electronic store front on some servers. And the point the article is making: that it’s cheap compared to the seeming “industry standard” 30% shows really that there is and has been a price cartel between the tiny amount of major players in the electronic market place.

This is the kind of monopoloy I have been talking about since the beginning of 2019.

China’s Space Station Is Closer to Reality With Launch of Core Module

Posted on by

China today launched the main module of its new space station into low Earth orbit. The ambitious project is set to be China’s answer to the International Space Station, which has never included China in its membership.

The 55-foot core module is called Tianhe, or Harmony of the Heavens. It blasted off from the Wenchang Launch Center in Hainan in the wee hours of Thursday morning, late Wednesday night for the United States. It launched aboard a 190-foot-tall Long March-5b Rocket, which has been the flagship launcher of the program since 2016. This is the first of 11 launches planned to see the finished product of the Chinese Space Station in operation by late 2022.

Should all go according to plan, Tianhe is the section of the station that will actually house Chinese astronauts, for stints of up to half a year. The next launches will send up two experimental modules, which will attach to either side of Tianhe, four cargo shipments, and four crewed missions, the first of which is slated for June. Tianhe has a total of five docking ports, which could be expanded to six.

The core module is the largest spacecraft yet developed by China, according to Chinese state media. The total station weight will be around 66 tons. While a far cry from the over 450 tons the ISS was at its completion in 1998, the main goals of the space station—conducting experiments in space and exploring how properties of space affect the results—doesn’t really require a ton of room.

[…]

Source: China’s Space Station Is Closer to Reality With Launch of Core Module

Stratolaunch sends world’s biggest airplane on second test flight

Posted on by

Stratolaunch, the aerospace company founded by the late Seattle billionaire Paul Allen, put the world’s biggest airplane through its second flight test today, two years after the first flight.

“We are airborne!” Stratolaunch reported in a tweet.

Today’s takeoff from California’s Mojave Air and Space Port at 7:28 a.m. PT marked the first time the plane, nicknamed Roc after the giant bird of Arabian and Persian mythology, got off the ground since Stratolaunch’s acquisition by Cerberus Capital Management in October 2019.

Roc rose as high as 14,000 feet and traveled at a top speed of 199 mph during a flight that lasted three hours and 14 minutes — which is close to an hour longer than the first flight on April 13, 2019. During that earlier flight, the airplane reached a maximum speed of 189 mph and maximum altitude of 17,000 feet.

Zachary Krevor, Stratolaunch’s chief operating officer, said today’s flight accomplished all of its test objectives by checking the performance of improved instrumentation, a more robust flight control system and an environmental control system that allowed the pilots to work in a pressurized cockpit. Krevor said the crew included chief pilot Evan Thomas, pilot Mark Giddings and flight engineer Jake Riley.

[…]

Since Roc’s first flight in 2019, the business model for the 10-year-old venture has shifted: In its early years, Stratolaunch focused on using Roc as a flying launch pad for sending rockets and their payloads to orbit. The concept capitalizes on the air launch system pioneered by SpaceShipOne, which won financial backing from Allen and won the $10 million Ansari X Prize in 2004,

The new owners still expect to use Roc for air launch, but the current focus is on using the plane as a testbed for Stratolaunch’s hypersonic flight vehicles, Once the plane is cleared for regular operations, perhaps next year, Stratolaunch could begin launching its Talon-A prototype hypersonic plane.

[…]

Other companies, principally including Virgin Orbit, are also working on next-generation air launch technology. Such systems hold the promise of greater versatility and quicker response time for launching payloads, due to the fact that the carrier planes can take off from a wide variety of runways, fly around inclement weather and theoretically launch their payloads in any desired orbital inclination.

Stratolaunch’s twin-fuselage, six-engine Roc airplane is in a class by itself, thanks to its world-record wingspan of 385 feet. In comparison, the wingspan of the modified Boeing 747 that Virgin Orbit is using comes to 211 feet. The previous record-holder was the Spruce Goose, a prototype seaplane that made its debut in 1947 and had a 320-foot wingspan. Built by Mojave-based Scaled Composites, Roc has the capacity to carry more than 500,000 pounds of payload.

Source: Stratolaunch sends world’s biggest airplane on second test flight

F-22 And F-35 Datalinks *Finally* Talk Freely With Each Other Thanks To A U-2 Flying Translator

Posted on by

Five F-35A Joint Strike Fighters and a single F-22 Raptor “talked” with each other using their proprietary stealthy datalinks via a U-2S Dragon Lady spy plane carrying a specialized communications gateway payload, during a recent demonstration. This marks the first time that the Air Force’s two stealth fighters were able to exchange data freely in flight, something that has been years in the making. The U-2 was also able to simultaneously share information with assets on the ground and at sea, as well as with non-stealthy combat aircraft, all in near-real-time. That info was used to initiate strikes from ground-based artillery and naval assets as part of the high-stakes capability demonstration.

This demonstration event was known as Project Hydra. The company’s Skunk Works advanced projects division worked together with the Air Force and the Missile Defense Agency (MDA) to carry out the tests. Elements of the U.S. Army and U.S. Navy were also involved.

[…]

Source: F-22 And F-35 Datalinks Finally Talk Freely With Each Other Thanks To A U-2 Flying Translator

It only took 10 years or so. For a military so entrenched in netcentric engagement to have their premier aviation assets not be able to communicate at all for so long is a major embarrassment. That they can only do it using a U2 within range is pretty weak.

Superspreaders of Malign and Subversive Information on COVID-19: Russian and Chinese Efforts Targeting the United States

Posted on by

Both Russia and China appear to have employed information manipulation during the COVID-19 pandemic in service to their respective global agendas. This report uses exploratory qualitative analysis to systematically describe the types of COVID-19-related malign and subversive information efforts with which Russia- and China-associated outlets appear to have targeted U.S. audiences from January 2020 to July 2020 and organizes them into a framework. This work lays the foundation for a better understanding of how and whether Russia and China might act and coordinate in the domain of malign and subversive information efforts in the future.

[..]

Key Findings

  • Both countries disseminated messages through a wide variety of channels and platforms, including social media.
  • Both countries attempted to tarnish the reputation of the United States by emphasizing challenges with its pandemic response and characterizing U.S. systems as inadequate.
  • Both countries falsely accused the United States of developing and intentionally spreading the virus.
  • The two countries appeared to differ in their principal goals for COVID-19-related information efforts: Russia aimed to destabilize the United States; China aimed to protect and enhance its own international reputation.
  • Both countries modified their COVID-19-related messaging over time, focusing on conspiracy theories about the virus’s origins and impacts from March 2020 to April 2020 and later moving to concentrate on perceived U.S. failure in responding to the pandemic.
  • While Russia deployed media with wide-ranging ideologies and a variety of audiences, China-linked messaging was ideologically uniform, consistent across multiple information outlets, and appeared to target audiences that were less varied.
  • Countering apparent Russian and Chinese malign and subversive information efforts will require campaigns that consider the capabilities and thematic emphasis of each of these actors.
  • Profiling Russian and Chinese sources known to frequently create and disseminate disinformation and propaganda can also inform counter-messaging efforts.
  • China and Russia appear to amplify one another’s messages, when opportune. This might eventually lead to some collaboration, albeit limited in nature.
  • Public health messaging should account for potential impacts of Russian and Chinese messaging on vaccination uptake

Source: Superspreaders of Malign and Subversive Information on COVID-19: Russian and Chinese Efforts Targeting the United States | RAND

Tesla Loses A Lot Of Money Selling Cars, But Makes It All Back On Credits And Bitcoin

Posted on by

On Monday after the close of business, Tesla announced its Q1 2021 financial results in its quarterly earnings call. The company turned a surprisingly large profit this quarter, but it didn’t do it by selling cars. Q1 net profit reached a new record for Tesla, at $438 million. Revenue for the electric car company was up massively to $10.39 billion. Unfortunately, all of that profit is accounted for in the company selling $518 million in regulatory credits, and $101 million was found in buying and then later selling Bitcoin.

That second point is particularly interesting, as Tesla purchased $1.5 billion worth of BTC, announced that the company would begin accepting BTC as payment for its cars, which drove up the value of BTC, then sold enough BTC to make a hundred million in profit. Strange how that works, eh? Surely nothing untoward going on there. Not at all. DOGE TO THE MOON! #hodlgang

Without the $619 million in credits and BTC sales, Tesla would have actually managed to lose $181 million in Q1. In that time, the company shifted 184,800 3/Y units, and while it didn’t build a single X or S in Q1, it sold 2020 units from previously-built inventory. That means the company lost around $970 per car sold in Q1.

[…]

Source: Tesla Loses A Lot Of Money Selling Cars, But Makes It All Back On Credits And Bitcoin