Documents made public by Heijdenrijk appear to show how a specialty firm helped Nawfal’s team attempt to boost his Twitter presence and gain traction on its audio Spaces with the use of “inorganic engagement.” In June, Heijdenrijk posted part of a pitch deck that detailed a plan to increase Nawfal’s Twitter engagement… Screenshots of a group WhatsApp chat posted by Heijdenrijk, along with additional documents and text messages viewed by NBC News, appear to show a discussion with representatives from a company called Winn.solutions. The messages include the company offering “growth hacking” services that guaranteed engagement from a “personal network of ‘super-fan’ accounts” for Nawfal’s accounts, using a mix of “inorganic” and “organic” engagement. ...

How is this legal? Insane. The $1 billion loan temporarily shifted a significant chunk of capital for SpaceX, which is intertwined with many of the most important space missions at the National Aeronautics and Space Administration and the Pentagon, to its top leader. … But I think I may have missed the main issue. The main issue is that Musk runs a web of companies, some of which need cash and some of which have cash, and he is also in the business of starting or buying other companies, creating more cash needs. If all of the companies are private, he can sort of use their cash interchangeably: Sure, each company has different outside investors and its own board of directors, but they are all part of the Elon Musk family, and it’s not too hard to work something out. Musk goes to the SpaceX board, he says “I need to borrow $1 billion to buy Twitter and save humanity,” they are like “your mission to save humanity is why we love you and why we are on the board, here you go,” and it’s fine. Probably he pays the loan back promptly with interest and everything is fine, but even if for some reason he couldn’t do that, probably SpaceX’s board and investors would understand. They love Musk and trust him completely. ...

“If he were to return,” Brent S. Isaacson, the campus police chief at the time, said in a July court filing, “the public’s disgust would extend to this campus, and we would again be viewed by many members of the public as sympathetic to Kershnar’s views and therefore at risk of violence.” … The case reflects continuing tensions over how universities should handle online conflagrations, freewheeling academic discourse and campus safety. Can public universities, which are bound by the First Amendment, restrict professors from campus because of comments they made on a podcast? Should they do so when threats are involved? And what is the marker of an actual threat, anyway? ...

In sum, Khan’s maximalist charge — that the entire no-confidence vote was orchestrated by Washington as part of a plot to shunt him from power — is, at least by the evidence released so far, an overstatement. But the US State Department did clearly make use of the existing no-confidence vote, which was born in the first place from Khan’s domestic failings and Pakistan’s internal political machinations, to lean on its government by making a mob boss–like threat, to ensure the vote went the way it wanted. ...

Lets hope this doesn’t take off :) Here is how it works: LinkedIn is using AI to generate questions about professional topics that will rank in the search engines. They are likely taking keywords that get traffic and then using all the other content on their platform to train a model that generates questions related to this keyword. They then use AI to generate the structure and headlines of the article. Additionally, they use AI to generate a paragraph of “starter content” under each headline. Now here comes the “magic.” They then put that article in the LinkedIn feed and prompt people to comment on each of the sections. This adds UGC content to the AI generated article. These articles then generate both acquisition and re-engagement. The articles start to rank in the search engines driving user acquisition and bringing back existing users to the platform. Via BrianBalfour.com/essays/a-breakdown-of-linkedins-ai-assisted-growth-loop

Fascinating, I didn’t notice this angle to content authors’ concerns before. Unlike authors who appear most concerned about retaining the option to remove their books from OpenAI’s training models, the Times has other concerns about AI tools like ChatGPT. NPR reported that a “top concern” is that ChatGPT could use The Times’ content to become a “competitor” by “creating text that answers questions based on the original reporting and writing of the paper’s staff.” ...

Great series of articles, love the forensic analysis but also the insight into the types of tests/validation that researchers should apply to their data to check it’s ‘sanity’ a four-part series detailing evidence of fraud in four academic papers co-authored by Harvard Business School Professor Francesca Gino Data Falsificada (Part 1) Data Falsificada (Part 2) Data Falsificada (Part 3) Data Falsificada (Part 4)

I remember taking tests on these type of networks in college but had never seen one in practice. After going reading through what it took, I think I was lucky :) Via Matt’s Tech Pages

…some in the industry fear the attitudes that created the conditions for the rapid growth and dramatic collapse of Celsius and FTX are still prevalent and that the industry remains vulnerable to influential individuals who are able to portray themselves as revolutionaries and pioneers. … The crypto industry “has an enormous attack vector for intelligent sociopaths,” says Travis Kling, cofounder of hedge fund Ikigai Asset Management. … Others have been left with an uncomfortable sense that crypto has not yet managed to clear its decks of bad actors and that, should another hype cycle arrive, the conditions that bred the likes of Celsius and FTX could recur. In short, that lessons have not necessarily been learned. ...