Microsoft has more layoffs, a new law to ban sale of cell phone location data and a security company demonstrates in real time how easy it is to poison a Large Language model.
 |
 |
 |
These and more top tech news stories on Hashtag Trending.
I’m your host Jim Love, CIO of IT World Canada and Tech News Day in the US.
Microsoft has confirmed another round of layoffs, the specifics of which remain undisclosed. This follows a significant reduction in January that saw the tech giant cut 10,000 jobs. The layoffs span across multiple teams, including sales, and come as part of a broader trend in the tech industry, with companies like Meta and Twitter also making significant cuts after a pandemic-induced hiring spree.
But these layoffs at Microsoft are a regular occurrence around the start of its new fiscal year on July 1. The company stated, “Organizational and workforce adjustments are a necessary and regular part of managing our business. We will continue to prioritize and invest in strategic growth areas for our future and in support of our customers and partners.”
Sources include: Axios
Paris-based startup Mithril Security has demonstrated the potential for poisoning large language models (LLMs), such as OpenAI’s ChatGPT, Google’s Bard, and Meta’s LLaMA, to spread misinformation.
The company manipulated an open-source model, GPT-J-6B, using the Rank-One Model Editing (ROME) algorithm, altering factual associations within the model. The tampered model was then posted on Hugging Face, an AI community website, under a misspelled repository name, demonstrating how unsuspecting developers could inadvertently incorporate a poisoned model into their applications.
Mithril’s demonstration underscores the potential for widespread dissemination of fake news through poisoned AI models. The company argues for the need for increased awareness and precautions by generative AI model users.
Mithril has an interest in getting attention for this issue – the company is developing an AICert service for cryptographically validating LLM provenance, emphasizing the importance of knowing the origin of AI models.
But Mithril’s demonstration shows how easily poisoned AI models could be distributed, even through reputable platforms like Hugging Face.
And it makes us think about the potential for targeted misinformation through poisoned AI models and raises concerns about the integrity of information generated by AI.
Sources include: The Register
A proposed law in Massachusetts, the Location Shield Act, aims to disrupt the data brokerage industry by prohibiting the sale of cellular location data. The law would require companies to obtain consent before collecting or processing such data.
Non-compliant companies could face legal action from the Massachusetts Attorney General’s office and potential class-action lawsuits. The legislation, a first of its kind in the U.S., was spurred by concerns over data privacy following the Supreme Court’s overturning of Roe v. Wade. Rights groups, including the ACLU and Planned Parenthood, support the Act as a means to protect individuals seeking medical procedures, such as abortions, from potential prosecution based on digital evidence.
The Location Shield Act, if passed, could set a significant precedent for digital privacy laws in the U.S., potentially influencing legislation in other states.
Sources include: Gizmodo
Meta’s new social media app, Threads, is set to introduce an alternative home feed of posts in response to user complaints.
Instagram boss Adam Mosseri confirmed that a feed for Threads, showing posts in chronological order from accounts users follow, is currently in the works.
This comes after users expressed frustration over the app’s algorithm choosing posts for them. Other planned features include post editing, language translation, and an easier way to switch between different Threads accounts. The company is also working on a desktop interface and a more robust search function.
The introduction of an alternative home feed in Threads shows how responsive Meta is trying to be to user feedback.
It shows just how aggressively the company is, in pursuing this window of opportunity to attract dissatisfied Twitter users. But it could set a precedent and expectations for how tech companies adapt their products based on user experience.
And that might not be such a bad thing.
Sources include: BBC
And finally, I caught a story in the Business Insider today about “The Tech Industry’s ‘Fake Work’ Problem: A Result of ‘Lazy Management’” according to the article.
The story talked about tech industry grappling with a growing issue of ‘fake work’, where employees are paid high salaries for work that contributes little to the company’s bottom line.
This phenomenon, which has been prevalent in tech giants like Amazon, Google, and Meta, was blamed on ‘lazy management’ rather than unproductive employees. Managers, in their pursuit of prestige and power, were tending to over hire, leading to bloated teams and a culture of ‘busywork’. This results in employees being assigned to projects that serve no mission-critical purpose or being left idle with no clear direction.
According to some, the big wave of recent tech layoffs was a result of aggressive hiring during the boom and then struggling to find meaningful work for the increased headcount during the bust.
And we might have thought of that as a blip – a pandemic issue, until that top story of Microsoft having a huge layoff that seems to be part of a management strategy.
I’m not picking on Microsoft, this is just a story we picked up today.
The issue is that these are not “headcount” or “bodies” – they are people and if you are going to manage with a blunt instrument that “takes them in and spits them out” – don’t be amazed when you find out how alienated and cynical your workforce can become.
There’s an old joke in Russia, that talks about management, saying, “they pretend to pay us, and we pretend to work.”
And we all know what that has done for creativity and innovation in Russia.
Just thinking out loud.
And that’s the top tech news stories for today.
Hashtag Trending goes to air five days a week with a special weekend interview episode called Hashtag Trending, the Weekend Edition. You can follow us on Google, Apple, Spotify or wherever you get your podcasts.
We’re also on YouTube five days a week with a video newscast only there we are called Tech News Day.
If you want to catch up on more news quickly, you can read these and more stories at TechNewsDay.com and at ITWorldCanada.com on the home page.
We love your comments. You can find me on Linked In, Twitter or on our Mastodon site and god forbid , I might even sign up for Threads.
Or just go to the article at itworldcanada.com/podcasts and you’ll find a text version with additional links and references. Click on the x if you didn’t like the stories, or the check mark if you did, and tell us what you think.
I’m your host, Jim Love. Have a Wonderful Wednesday!