OpenAI reportedly plans to charge up to $20,000 a month for specialized AI ‘agents’: $2,000/month for a "high-income knowledge worker" agent, $10,000/month for a software developer agent, and $20,000/month for an agent designed to support PhD-level research. OpenAI has secured a major commitment from SoftBank, which plans to spend $3B on these AI agents in 2025.
Microsoft unveils AI agents for sales, striking back at Salesforce: Microsoft launched Sales Agent and Sales Chat, AI-powered tools that automate lead qualification, scheduling, and customer insights. Salesforce CEO Marc Benioff has criticized Microsoft’s AI strategy, calling Microsoft a “reseller of OpenAI” and questioning the security risks of Copilot.
Google Search’s new ‘AI Mode’ lets users ask complex, multi-part questions: The new "AI Mode" allows users to ask complex, multi-part questions and follow up for deeper exploration, similar to Perplexity AI and ChatGPT Search. AI Mode uses a "query fan-out" technique to issue multiple related searches simultaneously for a more comprehensive response.
Amazon is working on a new 'reasoning' AI model that competes with OpenAI and Anthropic: Set to launch by June 2025 under the Nova brand, this model will combine quick answers with complex, multi-step reasoning.
DeepSeek Claims 545% Theoretical Profit Margin on AI Models: Chinese AI startup DeepSeek reported a theoretical 545% profit margin for its V3 and R1 models over a 24-hour period, based solely on inferencing costs like computing power, electricity, and data storage.
Apple’s Artificial Intelligence Efforts Reach a Make-or-Break Point: Apple Intelligence has rolled out slowly with underwhelming features like Genmoji and Image Playground, showing low real-world adoption despite claims of boosting iPhone sales. The next-gen Siri, aiming to rival ChatGPT and Alexa+, won’t launch until at least iOS 20 in 2027—five years behind competitors.
Deutsche Telekom & Perplexity to Launch AI-Powered Smartphone: The "AI Phone" will debut in 2026 for under $1,000, initially targeting the European market. DT and Perplexity are aiming to differentiate from OpenAI, Google, and Anthropic by embedding AI directly into a phone’s core experience.
OpenAI Introduces a $50M Consortium for AI Research and Education: The consortium includes prominent universities such as Harvard, MIT, Oxford, and Howard University, as well as organizations like Boston Children’s Hospital and the Boston Public Library.
Turnitin introduces Turnitin Clarity: A new tool designed to enhance academic integrity and transparency in student writing, particularly in the era of AI-assisted work. Educators can track students' writing processes, including typing patterns, pasted text, draft history, and AI tool usage.
‘I want him to be prepared’: why parents are teaching their gen Alpha kids to use AI: Parents, including computer science experts, believe AI proficiency is as important as coding for their children’s success. Unlike social media, which exposes kids to unfiltered content, parents see AI as a safer, more constructive tool when used responsibly.
AI Use in the Workplace: A majority of U.S. workers (63%) report little to no AI use at work, while only 1 in 6 use it regularly. AI users tend to be more educated, with over half holding at least a bachelor’s degree, and 22% having a postgraduate degree.
How AI Tools Are Reshaping the Coding Workforce: The role of developers is shifting from writing code to prompting and guiding AI tools effectively. Hiring has become more selective—companies no longer rush to fill seats but seek problem-solving and communication skills over just coding ability.
AI Will Upend a Basic Assumption About How Companies Are Organized: The economy is built on the idea that expertise is scarce and expensive. AI is about to make it abundant and practically free.
The printing press and mass literacy lowered the cost of knowledge, fueling economic transformations like industrialization. The internet further democratized information, and AI now represents the next step, making high-level expertise instantly accessible.
Organizations must rethink how they structure teams and allocate human talent when AI can handle both routine and high-level cognitive tasks.
Google co-founder Larry Page has reportedly formed a new AI company: The company, called Dynatomics, aims to use large language models to “create highly optimized designs for a wide variety of objects and then have a factory build them”.
A quarter of startups in YC’s current cohort have codebases that are almost entirely AI-generated: YC leaders, including CEO Garry Tan, believe AI-powered coding will become the dominant method of development moving forward.
Stealth Interview: Provides an undetectable AI-powered assistance tool designed to help job seekers cheat on technical interviews.
Sesame: AI voice model designed to deliver natural, expressive, and highly realistic speech. It aims to achieve "voice presence”.
Latitude: Build self-improving AI agents. Design, evaluate, and deploy truly autonomous AI agents integrated with your tools and data.
Societies: AI simulation of your LinkedIn network that helps users predict, refine, and enhance their LinkedIn posts before publishing. Runs thousands of AI-driven tests to estimate how many reactions a post will receive.
Turing Award Goes to 2 Pioneers of Artificial Intelligence: Andrew Barto and Richard Sutton received the Turing Award for their pioneering work in reinforcement learning (RL), a technique inspired by how humans and animals learn through trial and error by optimizing for rewards (pleasure) and avoiding negative outcomes (pain). Their research, which began in the 1970s and 1980s, became the foundation for modern AI breakthroughs like AlphaGo and ChatGPT.
AI companies race to use ‘distillation’ to produce cheaper models: Distillation is a technique where a large AI model (teacher) generates data to train a smaller, more efficient model (student) while retaining much of the teacher’s intelligence.
The Differences between Deep Research, Deep Research, and Deep Research
Anthropic’s Claude AI is playing Pokémon: Claude displays its thought process in a side window, allowing viewers to follow its logic and strategy. Unlike earlier versions, Claude 3.7 has defeated three Gym Leaders, showing improved adaptive and agentic capabilities.
Mistral OCR: Unlike traditional OCR tools, Mistral OCR can process not only text but also tables, images, equations (e.g., LaTeX), and handwritten notes. It processes up to 2,000 pages per minute on a single node
The Government Knows A.G.I. Is Coming: Experts now believe that AGI may emerge in as little as 2–3 years, much sooner than earlier estimates of 5–15 years. Traditional government structures are too slow to harness or regulate emerging AI. The U.S. Artificial Intelligence Safety Institute seeks to bridge public-private gaps, but major reforms are still needed.
Eric Schmidt argues against a ‘Manhattan Project for AGI’: The policy paper “Superintelligence Strategy” argues that a U.S.-led Manhattan Project for AGI could trigger hostile retaliation from China, including cyberattacks, destabilizing global relations. Instead of aggressively pursuing AGI dominance, the authors propose "Mutual Assured AI Malfunction" (MAIM)—disabling threatening AI projects preemptively via cyberattacks.
Scale AI is being investigated by the US Department of Labor: The DOL has been investigating Scale AI since at least August 2024 for possible misclassification of workers, unpaid wages, and illegal retaliation. A 2023 Washington Post investigation raised concerns about low wages for overseas contractors, though Scale AI claimed pay rates were improving.
Google and Amazon AI Say Hitler’s Mein Kampf Is ‘a True Work of Art’: AI struggles to differentiate analysis from endorsement, leading to misleading summaries, while training on AI-generated content further distorts information quality—deepening the crisis in search reliability.
The author of SB 1047 introduces a new AI bill in California: Wiener introduced SB 53, focusing on AI whistleblower protections and a public cloud computing cluster (CalCompute) for AI research. SB 53 protects employees who report AI systems posing a "critical risk", defined as potential harm leading to 100+ deaths or $1 billion+ in damages.
Judges Are Fed up With Lawyers Using AI That Hallucinate Court Cases: Multiple attorneys have been sanctioned for submitting AI-generated, fictitious legal citations, underscoring the dangers of unverified AI use in law.
Microsoft’s $13 Billion OpenAI Investment Cleared by UK Watchdog
Anthropic quietly removes Biden-era AI policy commitments from its website: Both OpenAI and Anthropic are actively seeking government contracts, potentially influencing their decisions to align with shifting federal policies.
OpenAI’s ex-policy lead criticizes the company for ‘rewriting’ its AI safety history: OpenAI now portrays AGI as a "continuous path," downplaying past caution over GPT-2. Brundage warns this could set a precedent for dismissing AI risks unless there's "overwhelming evidence" of harm, risking reckless deployment.
New Data Shows Just How Badly OpenAI And Perplexity Are Screwing Over Publishers: Despite claims from AI companies that their search tools increase publisher traffic, AI-driven search engines send 95.7% fewer click-throughs compared to traditional Google search. AI bot scraping increased by 117% from Q3 to Q4 2024, meaning AI models are consuming publisher content at a much higher rate. AI applications provide only a 0.37% referral rate when they scrape content—96% lower than Google search.
Pinterest Is Being Strangled by AI Slop: Many posts link back to AI-powered content farms, using fake bloggers and AI-generated bios to drive traffic for ad revenue. Pinterest users struggle to find real human-made content, with some calling the platform “dead” due to AI spam.
Google’s Sergey Brin Urges Workers to the Office ‘at Least’ Every Weekday: Brin believes Google has all the necessary components to win the race to AGI but needs to “turbocharge” efforts.
Christie's AI art auction outpaces expectations, bringing in more than $728,000
No posts
Comments
Nothing yet. Say the first thing.
Sign in to join the conversation.