Issue 53 21st May 2024

Highlights

This Week in AI: OpenAI moves away from safety

Keeping up with an industry as fast-moving as AI is a tall order. So until an AI can do it for you, here’s a handy roundup of recent stories in the world of machine learning, along with notable research and experiments we didn’t cover on their own. By the way, TechCrunch plans to launch an AI newsletter […]

Hello GPT-4o

We’re announcing GPT-4o, our new flagship model that can reason across audio, vision, and text in real time.

As the AI world gathers in Seoul, can an accelerating industry balance progress against safety?

Companies such as OpenAI and Meta push ahead, but it is clear that biggest changes are yet to come This week, artificial intelligence caught up with the future – or at least Hollywood’s idea of it from a decade ago. “It feels like AI from the movies,” wrote the OpenAI chief executive, Sam Altman, of his latest system, an impressive virtual assistant. To underline his point he posted a single word on X – “her” – referring to the 2013 film starring Joaquin Phoenix as a man who falls in love with a futuristic version of Siri or Alexa, voiced by Scarlett Johansson. Continue reading...

Achieving Sustainability Goals with PyTorch and Intel AI

This post was contributed by Intel AI in partnership with the PyTorch Foundation.

Video

OpenAI Spring Update - Recap

We are looking forward to hearing what they are bringing. Join Sammy and Gavita afterwards to discuss the new developments!

Articles

Mastering GenAI ML System Design Interview: Principles & Solution Outline

Strategies and Insights from Both Sides of the Interview Table.

How to Evaluate Your Predictions

As a general rule: One can use MSE to evaluate mean predictions, MAE to evaluate median predictions, the quantile score to evaluate more general quantile predictions and the energy or MMD score to evaluate distributional predictions. Now for a new test point y, we want to score your prediction, that is you want a function S(y,f( x )), that is minimized (in expectation) when f( x ) is the best thing you can do.

How to Setup a Multi-GPU Linux Machine for Deep Learning in 2024

Set up a Multi-GPU Linux system with necessary libraries such as CUDA Toolkit and PyTorch to get started with Deep Learning . We will install 1) CUDA Toolkit, 2) PyTorch and 3) Miniconda to get started with Deep Learning using frameworks such as exllamaV2 and torchtune.

Spam, junk … slop? The latest wave of AI behind the ‘zombie internet’

I'm quoted in this piece in the Guardian about slop: I think having a name for this is really important, because it gives people a concise way to talk about the problem. Before the term ‘spam’ entered general use it wasn’t necessarily clear to everyone that unwanted marketing messages were a bad way to behave. I’m hoping ‘slop’ has the same impact – it can make it clear to people that generating and publishing unreviewed AI-generated content is bad behaviour.

The Surprising Benefit Of Gen AI Technology

Generative AI is yielding a surprising benefit for productivity in business operations. How can companies embrace and ramp up this benefit? And what are the implications for third-party services?

GPT 4o is both a God-send, and a Busted Flush — part 2

The saga continues, but this time with a slightly better outcome, Maybe. Continue reading on Medium »

A Complete Guide to BERT with Code

Overview: The encoder block in the Transformer is responsible for accepting an input sequence, and creating rich, numeric vector representations for each word (or more specifically, each token). The original BERT model was released shortly after OpenAI’s Generative Pre-trained Transformer (GPT), with both building on the work of the Transformer architecture proposed the year prior.

Google remains focused on its long quest for your eyeballs

AI Overviews, announced this week, are the culmination of a long line of products dedicated to keeping you on Google.com Google announced this week that it would begin the international rollout of its new artificial intelligence-powered search feature, called AI Overviews. When billions of people search a range of topics from news to recipes to general knowledge questions, what they see first will now be an AI-generated summary. Google touted AI Overviews at its annual I/O developer conference as a way of delivering customers quick answers and simplifying the online search experience, but it also has another effect on the way that people engage with the internet: keeping users, and advertisers, on Google.com. It’s a new era in Google’s years-long quest for your attention. Continue reading...

Introducing depyf: mastering torch.compile with ease

is a cornerstone of PyTorch 2.x, offering a straightforward path to accelerate machine learning workflows with just a single line of code for both training and inference. We are thrilled to introduce, a new project to the PyTorch ecosystem designed to help users understand, learn, and adapt to !

Upcoming events

Microsoft – Watch the May 20 event

Tune in for the Microsoft Event on May 20 at 2:00pm PST to hear an announcement from Microsoft.