Lifeboat Foundation

In every downturn, we tend to measure the pain by counting layoffs. (Dell is the latest, announcing it will cut 6,650 jobs or 5% of its workforce.) According to Layoffs.fyi, a smart if incomplete tracker of job cuts, tech companies laid off almost 95,000 workers in the first five weeks of this year, which is already about 60% of the layoffs it reported for all of 2022.

While job cuts are normal, there’s something different about this economic dip. To start, as Jena McGregor reports, the advent of remote work has cemented the digital pink slip.

Continue reading “How ChatGPT, Bard And AI Rivals Are Shaping Layoffs And Hiring” »

In the ever-evolving landscape of test preparation, a new player has sprouted on the scene – artificial intelligence.

At the forefront of this movement is a Korean start-up, Riiid, founded by YJ Jang, a graduate of the Haas School of Business at the University of California, Berkeley. Riiid has already made a name for itself in the Asian test-prep market for the TOEIC, a measure of English proficiency in the business world. Now, the company has set its sights on the American market with an SAT and ACT prep system called R.Test.

A.I. technology, with its mimicry of the networks of neurons in the human brain, has the potential to revolutionize the way educators approach their craft.

Though Bill Gates recently declared the current moment in technology as important as the advent of the PC, storied Silicon Valley firm Benchmark is taking a relatively cautious view of AI’s gold rush for now. A “big majority” of startups pitching to partner Chetan Puttagunta claim to be working on machine learning, but, he told Forbes.

Benchmark is leading the investment with a bet that MindsDB can replace the need to hire hundreds of machine learning engineers.

If your work involves analyzing and reporting on data, then it’s understandable that you might feel a bit concerned by the rapid advances being made by artificial intelligence (AI). In particular, the viral ChatGPT.

AI, particularly ChatGPT, has raised job security concerns among data analysts. Here we look at the potential impact and discuss that despite limitations like frequent mistakes and limited data upload capabilities, ChatGPT has the potential to automate data gathering and analysis tasks in the future.

After six years of peace, the two tech giants are on course to butt heads again over the future of artificial intelligence.

Microsoft is about to go head-to-head with Google in a battle for the future of search. At a press event later today, Microsoft is widely expected to detail plans to bring OpenAI’s ChatGPT chatbot to its Bing search engine. Google has already tried to preempt the news, making a rushed announcement yesterday to introduce Bard, its rival to ChatGPT, and promising more details on its AI future in a press event on Wednesday.

The two tech giants are on course to butt heads again.

Setting boundaries for these tools, then, could be crucial, some researchers say. Edwards suggests that existing laws on discrimination and bias (as well as planned regulation of dangerous uses of AI) will help to keep the use of LLMs honest, transparent and fair. “There’s loads of law out there,” she says, “and it’s just a matter of applying it or tweaking it very slightly.”

At the same time, there is a push for LLM use to be transparently disclosed. Scholarly publishers (including the publisher of Nature) have said that scientists should disclose the use of LLMs in research papers (see also Nature 613, 612; 2023); and teachers have said they expect similar behaviour from their students. The journal Science has gone further, saying that no text generated by ChatGPT or any other AI tool can be used in a paper⁵.

One key technical question is whether AI-generated content can be spotted easily. Many researchers are working on this, with the central idea to use LLMs themselves to spot the output of AI-created text.

Diffusion models have recently produced outstanding results on various generating tasks, including the creation of images, 3D point clouds, and molecular conformers. Ito stochastic differential equations (SDE) are a unified framework that can incorporate these models. The models acquire knowledge of time-dependent score fields through score-matching, which later directs the reverse SDE during generative sampling. Variance-exploding (VE) and variance-preserving (VP) SDE are common diffusion models. EDM offers the finest performance to date by expanding on these compositions. The existing training method for diffusion models can still be enhanced, despite achieving outstanding empirical results.

The Stable Target Field (STF) objective is a generalized variation of the denoising score-matching objective. Particularly, the high volatility of the denoising score matching (DSM) objective’s training targets can result in subpar performance. They divide the score field into three regimes to comprehend the cause of this volatility better. According to their investigation, the phenomenon mostly occurs in the intermediate regime, defined by various modes or data points having a similar impact on the scores. In other words, under this regime, it is still being determined where the noisy samples produced throughout the forward process originated. Figure 1(a) illustrates the differences between the DSM and their proposed STF objectives.

Figure 1: Examples of the DSM objective’s and our suggested STF objective’s contrasts.

It’s a 40-minute livestream.

🧚🐝👾

“I’ll teach you how to jump on the wind’s back, and then away we go,” Peter Pan says to Wendy.

In J.M. Barrie’s book, fairies can be brought back to life if enough people believe in them.

Continue reading “Clap if you believe in robot fairies” »