Computerspeak by Alexandru Voica

During a media tour to hype up Grok 3 ahead of its release, Elon Musk appeared at Welt’s Economic Forum in Germany where he put superintelligence among the top 10 most important milestones in the history of life on Earth. I explain the consequences of his statement in this week's newsletter: https://www.computerspeak.co/p/agi-making-humanity-multiplanetary

The most meaningful achievement of the AI Action Summit in Paris is not what happened at the event, but in fact what we didn’t see. For the first time at a global AI event, the AI grift industrial complex, comprised of alarmists, doomers, skeptics, privacy and safety fanatics, and other conspiracy theorists, was relegated to where they’ve always belonged: at fringe events, preaching to their cleve...

In recent months, a new breed of reasoning models has begun to emerge from the research labs of the tech world. An early adopter of these reasoning models has been the software engineering community. Instead of expecting the models to autonomously produce flawless code, engineers are leveraging them as interactive partners—similar to having a knowledgeable colleague or an advanced debugging tool....

When Chinese AI research lab DeepSeek released R1, an open weights AI model that was on par with OpenAI’s o1 reasoning model, the response from many in the American media and industry circles was swift and alarmist. The achievement was quickly framed as a "Sputnik moment"—a term evoking the Soviet Union’s launch of the first artificial satellite in 1957 and the subsequent Cold War space...

During a press conference held on Tuesday at the White House, Donald Trump, together with Larry Ellison from Oracle, Masayoshi Son from SoftBank and Sam Altman from OpenAI, announced the Stargate Project, a new company set to invest $500 billion over the next four years to build AI infrastructure in the United States. So what exactly does that get you in terms of real-world infrastructure? We do t...

Content moderation at scale remains one of the most complex challenges that technology companies have to deal with today. I’ve seen firsthand how different companies approach this challenge across different platforms, from content creation to distribution. Learn more at https://www.computerspeak.co/p/dealing-with-precision-in-content-moderation

An ambitious new AI model called GET (General Expression Transformer) promises to rewrite the rulebook for transcriptional regulation, a cornerstone of biological processes. Unveiled by researchers from MBZUAI, Columbia University and Carnegie Mellon, GET is not just another computational biology tool—it’s a leap forward in the form of a foundational model designed to predict gene expression acros...

A new report from MMC Ventures discusses the ethical challenges of training AI models, drawing parallels between the data gathering practices of today and the Napster-to-Spotify transition that occurred over two decades ago in the music industry. The report examines the perspectives of three stakeholders: content creators, data rights holders, and AI developers, highlighting the need for fair comp...

Traditionally, we used to measure companies by their multiple, calculated as the company's valuation divided by its revenue. Multiples used to be a safe metric for investors to compare a company to other businesses. But with the AI companies of today, using multiples is no longer an option because, for many of them, they create a distorted picture of their true worth. Learn more at https://www.com...

A new paper presented this week by researchers from MBZUAI and UC Berkeley at the EMNLP conference in Miami explains why Romanian or Cherokee (a language spoken by 2,000 people in the United States) can both be considered low resource, albeit for different reasons. Learn more at https://www.computerspeak.co/p/better-definition-low-resource-languages Also in the news this week: Wired: Inside the Bi...

TikTok’s parent company ByteDance has just previewed  X-Portrait 2 , a new portrait animation model built in cooperation with researchers from Tsinghua University, China’s best-ranked university in science and technology. The model is an evolution of the first-generation X-Portrait technology (which was presented earlier this year at SIGGRAPH) and provides similar functionality to  Runway’s Act-On...

Welcome to Computerspeak! I’m starting this newsletter and podcast to keep track of new breakthroughs in AI and other emerging technologies, and share them with you every week. Computerspeak will be published on Friday on a weekly basis, providing a selection of the most interesting media articles about AI. Learn more at https://www.computerspeak.co/© 2024 Black Knight Entertainment Ltd

Countries in Asia are adopting AI at an accelerated rate and, if that trend continues, we’re likely to see AI-related investments in the region surpass $100 billion by 2028, according to IDC. Learn more at https://www.computerspeak.co/ © 2024 Black Knight Entertainment Ltd

Several reporters have been telling me that there's mounting skepticism of AI startups and their revenue potential among investors in the Valley. Here are the are two contributing factors to this cooling off. Learn more at https://www.computerspeak.co/ © 2024 Black Knight Entertainment Ltd

My thoughts after attending TED AI Vienna, the first AI-focused TED conference held in Europe. Learn more at https://www.computerspeak.co/ © 2024 Black Knight Entertainment Ltd

Remember nine months ago when the world collectively was rolling its eyes every time Mark Zuckerberg mentioned the metaverse on an earnings call? Well, maybe it’s time to roll them back, put on a pair of Ray-Ban Meta glasses, and take a closer look at what Meta is doing because the metaverse could be finally revealing itself to us in its hype-less, tangible future glory. Learn more at https://www....

In the race for AI supremacy, a new innovation hub is emerging according to a story from Reed Albergotti in Semafor: the Middle East. As the UAE and Saudi Arabia build data centers capable of training foundational models, more startup founders and investors are relocating to the region, taking advantage of recently established free zones where entrepreneurs can experiment freely with emerging tech...

In her new book Supremacy, Parmy Olson goes behind the scenes of the AI industry’s two-year rollercoaster ride which began in November 2022 with the launch of ChatGPT. She focuses on the evolution of two founders: Sam Altman of OpenAI and Demis Hassabis of Google DeepMind, and chronicles their journeys from the early startup days where both had lofty goals of solving science and providing economic...

The Oxford Internet Institute has published a paper highlighting the uneven distribution of computing around the world, identifying three categories of countries when it comes to access to cloud AI infrastructure. Learn more at https://www.computerspeak.co/ © 2024 Black Knight Entertainment Ltd

According to some commentators, we’ve reached peak AI and the industry is in a huge bubble because generative AI provides no real benefits to early adopters. Here is their argument: over the last year, Big Tech has made sizable infrastructure investments in AI and VCs have signed large checks hundreds of startups (some with eye-watering valuations) but, so far, we have nothing to show for it. Lear...

For years, Apple has perfected the art of making technology accessible to the point of it becoming invisible—if you’re using an iPhone or a Mac, everything just works. They offer a similar vision for AI with Apple Intelligence which will improve apps and OS features in small yet meaningful ways. Learn more at https://www.computerspeak.co/ © 2024 Black Knight Entertainment Ltd

Shortly after appearing on Dwarkesh Patel’s podcast, Leopold Aschenbrenner, a former OpenAI safety researcher, posted a chart on X arguing that, given the current pace of progress with AI, we will have AGI by 2027 and maybe even “superintelligence” beyond 2030 thanks to millions of GPUs concentrated in 10 GW data centers. Learn more at https://www.computerspeak.co/ © 2024 Black Knight Entertainmen...

The generative AI race welcomed another important contender last week when Meta released two pre-trained and fine-tuned Llama 3 language models with 8B and 70B parameters, respectively. Just in the past three months, Microsoft, Google, Databricks, Cohere and Mistral have all made foundational models available for broad use, creating a growing trend among technology companies. Learn more at https:/...

Consumer AI devices are having a difficult time lately. With global shipments of smart home devices declining in 2023, it seemed only natural that the market would look to generative AI to breathe new life into existing devices or expand with a host of new, mobile AI-powered gadgets. However, a different story is playing out, with a string of recent releases failing to capitalize on the potential...

Ever since Kodak introduced the first camera for amateur photographers in 1888, people have preserved precious memories by carefully curating photo albums—initially as physical books and more recently as digital collections stored in the cloud with services such as Google Photos, Dropbox or Flickr. But thanks to rapid advancements in generative AI, the century-old concept of a photo album is about...