Welcome to this week’s edition of TekTonic. OpenAI’s unveiling of the “o1-preview” and “o1-mini” models continues reverberating, moving beyond predictive text to decision-based reasoning—a leap that could redefine AI’s potential. Companies are rethinking strategies amid AI’s rapid evolution, acknowledging its transformative power and impact on the workforce, such as developing in-house AI solutions to stay competitive. Globally, efforts intensify to regulate and harness AI responsibly, reflecting heightened awareness of its far-reaching implications. From innovations in AI hardware enhancing cloud capabilities to concerns over data authenticity and user privacy, this week’s stories highlight dynamic changes influencing industries and daily life alike.
This week’s biggest story is undoubtedly the revelations of OpenAI’s new model, “o1,” released last week, representing a fundamentally different “species” of the language model. The implications of this development will propel AI faster and further than we previously thought possible. Unlike autoregressive models like GPT-1, 2, 3, and 4—whose main goal was to predict the next word (system 1 thinking)—”o1″ transcends this paradigm by engaging in system 2 thinking, focusing on predicting the best decision.
The available training data limited models like GPT-4 to improve performance—a resource running out. However, system two models like “o1” don’t require vast data. Instead, they rely on the quality of the training data and the effectiveness of the reasoning derived from it. Remember that “o1” was trained for reasoning and that reasoning was not checked by humans but by the model’s twin. Because of this, human limitations do not constrain this system’s growth. As evidence, IQ tests administered to the “o1-preview” show it performs at 100 points (50%), increasing to 120 (90%) when it can reference its training data. In light of this, Sam Altman’s recent comments that “o1” is like GPT-2 make sense.
We begin with OpenAI’s latest unveiling of its new reasoning model, GPT-o1. Touted by experts as a significant step toward artificial general intelligence, this model showcases exceptional capabilities in coding and problem-solving. Reports suggest that GPT-o1 can pass OpenAI’s own engineering hiring tests with a 90-100% success rate. The model’s prowess has even sparked debates about the possible obsolescence of coding as a human profession, highlighting the accelerating pace at which AI is evolving.
In the corporate arena, Salesforce CEO Marc Benioff announced its new AI strategy while acknowledging that its AI integration may lead to job displacement. This marks one of the first times a major tech company has openly recognized AI’s potential impact on employment. In a turn of events, Klarna’s CEO plans to cancel its Salesforce subscription to build its own version in-house with the help of GenAI, to which the Salesforce CEO responded, “Nah.”
Meanwhile, T-Mobile has entered a multi-year partnership with OpenAI to develop IntentCX, an AI-driven customer experience platform. Companies are now harvesting customer intent rather than sentiment or product preferences.
As AI technologies become more sophisticated, leading scientists and AI pioneers from the United States, China, and other nations have called for international protections against potential “catastrophic risks.” They advocate for establishing a global authority to oversee artificial intelligence development and deployment.
In response to growing concerns over deepfakes and misinformation, Google is advancing AI image authenticity by integrating the Coalition for Content Provenance and Authenticity (C2PA) standard into its search features.
In hardware news, Intel is developing new AI-capable chips for Amazon’s cloud computing unit in a multibillion-dollar deal. These chips are expected to enhance Amazon Web Services’ ability to provide robust AI solutions to its clients.
Advanced AI tools are also becoming accessible. Companies like Mistral AI have launched free tiers for developers to test powerful AI models, while productivity platforms such as Slack also integrate AI agents from partners like Asana, Adobe, and Workday. These integrations aim to streamline workflows and enhance collaboration.
In healthcare and biotechnology-related AI news, Elon Musk’s company, Neuralink, has received approval from the U.S. Food and Drug Administration for its experimental brain implant, Blindsight. This revolutionary device aims to restore vision to the blind and assist individuals with paralysis by directly interfacing with the brain.
As artificial intelligence becomes increasingly integrated into daily life, governments worldwide are increasing efforts to regulate its development and use. Malaysia has launched its National Guidelines on AI Governance and Ethics. Similarly, Hong Kong is preparing to release AI policies for the financial sector by the end of October 2024.
In the United States, a new collaboration between a state government and Google is set to revolutionize the administration of unemployment benefits. Google’s generative AI system will analyze transcripts from appeals hearings, and issue recommended decisions, aiming to clear a backlog of claims efficiently.
Turning to the business sector, a significant shift is occurring within the gaming industry. From Blizzard to Bethesda, game developers are increasingly forming unions, with the Communications Workers of America playing a pivotal role in this movement. Workers are organizing to address workplace conditions, job security, fair compensation, and artificial intelligence.
In consumer technology news, Apple is experiencing a slump in pre-orders for its latest products. Analysts attribute the slump to delays in implementing advanced AI features called “Apple Intelligence.” Consumers seem to be holding off on upgrading as they await these anticipated AI capabilities. Maybe Apple is embracing the new tech mantra “over promise and under deliver”.
In media and entertainment, the U.S. Federal Trade Commission has released a report expressing concerns about social media users’ lack of control over how artificial intelligence systems utilize their data. The report emphasizes that companies collect, share, and process vast amounts of user information with little transparency, especially regarding AI-driven content curation and targeted advertising.
The intersection of AI and entertainment is further exemplified by a new partnership between AI video startup Runway and Lionsgate, the studio behind the “John Wick” franchise. Runway will use Lionsgate’s extensive catalog to train its AI models, which filmmakers can leverage to create new content.
In a surprising move blending politics and finance, former President Donald Trump has announced the launch of a new cryptocurrency venture, World Liberty Financial. Alongside his sons and crypto entrepreneurs, the platform aims to promote stablecoins and potentially serve as a borrowing and lending platform.
In education news, Sam Altman, CEO of OpenAI, addressed high school students, sharing his perspective that dropping out of college was not a significant setback in his journey. He emphasized that traditional educational paths are not the only route to success in the rapidly evolving tech industry.
That is it. Stay tuned for more news next week. I APPRECIATE ALL THE READERS. Help spread the word. I put these newsletters together so productive humans like you can stay on top of the latest AI stories shaping our world.
If you like these weekly tech news reports, subscribe to get notified of new editions and updates. For daily updates, check out our news page. For a more in-depth analysis of the week’s news, sign up for our free weekly newsletter to the right of the daily news, or follow me on Twitter or YouTube.
