home.social

#ai-agent — Public Fediverse posts

Live and recent posts from across the Fediverse tagged #ai-agent, aggregated by home.social.

fetched live
  1. MIT researchers have developed GIFT, a system that teaches vision-language AI models to generate accurate CAD programmes for 3D objects. The method is more precise than existing techniques while using a fraction of the computation, enabling faster rapid prototyping. news.mit.edu/2026/turning-2d-d #AIagent #AI #GenAI #AIResearch

  2. MIT researchers have developed GIFT, a system that teaches vision-language AI models to generate accurate CAD programmes for 3D objects. The method is more precise than existing techniques while using a fraction of the computation, enabling faster rapid prototyping. news.mit.edu/2026/turning-2d-d #AIagent #AI #GenAI #AIResearch

  3. SpaceXAI has open-sourced Grok Build, the Rust-based agent harness and terminal interface behind its grok coding CLI. The Apache 2.0 licensed release covers the agent loop, tool dispatch and extension system. Grok 4.5 itself remains closed. marktechpost.com/2026/07/15/sp #AIagent #AI #GenAI #AgenticAI

  4. SpaceXAI has open-sourced Grok Build, the Rust-based agent harness and terminal interface behind its grok coding CLI. The Apache 2.0 licensed release covers the agent loop, tool dispatch and extension system. Grok 4.5 itself remains closed. marktechpost.com/2026/07/15/sp #AIagent #AI #GenAI #AgenticAI

  5. Applied Computing has raised 20M USD in Series A funding to build a foundation AI model for the entire oil, gas and petrochemical industry. The startup aims to deploy AI across entire plant operations, from drilling to refining. techcrunch.com/2026/07/15/appl #AIagent #AI #GenAI #EnterpriseAI

  6. Applied Computing has raised 20M USD in Series A funding to build a foundation AI model for the entire oil, gas and petrochemical industry. The startup aims to deploy AI across entire plant operations, from drilling to refining. techcrunch.com/2026/07/15/appl #AIagent #AI #GenAI #EnterpriseAI

  7. MIT researchers have developed 'neural transparency' - a tool that lets users peek inside AI neural networks before their chatbot ever speaks. The study found people consistently misjudge how their personalised AI will behave, overestimating good traits and underestimating harmful ones like sycophancy. news.mit.edu/2026/3-questions- #AIagent #AI #GenAI #AIEthics

  8. MIT researchers have developed 'neural transparency' - a tool that lets users peek inside AI neural networks before their chatbot ever speaks. The study found people consistently misjudge how their personalised AI will behave, overestimating good traits and underestimating harmful ones like sycophancy. news.mit.edu/2026/3-questions- #AIagent #AI #GenAI #AIEthics

  9. A survey of 101 enterprises finds that 71% admit most of their deployed AI agents are actually just chatbot wrappers, not true multi-step orchestrated workflows. Anthropic Claude leads platform adoption at 40%, but genuine agentic deployment remains the exception rather than the norm. venturebeat.com/ai/agentic-orc #AIagent #AI #GenAI #AIEnterprise

  10. A survey of 101 enterprises finds that 71% admit most of their deployed AI agents are actually just chatbot wrappers, not true multi-step orchestrated workflows. Anthropic Claude leads platform adoption at 40%, but genuine agentic deployment remains the exception rather than the norm. venturebeat.com/ai/agentic-orc #AIagent #AI #GenAI #AIEnterprise

  11. 🚀 Fastest-growing AI projects today

    1. The growth trends highlight an increasing demand for intelligent tools that integrate s...
    2. The **isjiamu/gzh-design-skill** repository a standout thweek, boasting a Growth Score...
    3. Thtool converts Markdown content into HTML suitable for WeChat's article editor, offeri...

    Full report → pullrepo.com/report/todays-ai-

    #AI #OpenSource #GitHub #Tech #AIAgent

  12. 🚀 Fastest-growing AI projects today

    1. The growth trends highlight an increasing demand for intelligent tools that integrate s...
    2. The **isjiamu/gzh-design-skill** repository a standout thweek, boasting a Growth Score...
    3. Thtool converts Markdown content into HTML suitable for WeChat's article editor, offeri...

    Full report → pullrepo.com/report/todays-ai-

    #AI #OpenSource #GitHub #Tech #AIAgent

  13. OpenAI has released a 230 USD keyboard called Codex Micro, designed with Work Louder. The specialised mini-keyboard features six frosted keys that display real-time feedback on Codex agent threads - showing white when idle, blue when thinking, green when complete, amber when human input is needed, and red for errors. techcrunch.com/2026/07/15/amid #AIagent #AI #GenAI #Agents

  14. OpenAI has released a 230 USD keyboard called Codex Micro, designed with Work Louder. The specialised mini-keyboard features six frosted keys that display real-time feedback on Codex agent threads - showing white when idle, blue when thinking, green when complete, amber when human input is needed, and red for errors. techcrunch.com/2026/07/15/amid #AIagent #AI #GenAI #Agents

  15. Vint Cerf, the father of the internet, has joined Innovation Labs to develop standards for AI agents to identify themselves online. The DNSid system would link autonomous agents to domain names using cryptographic proofs as agents spread across the internet. techcrunch.com/2026/07/15/vint #AIagent #AI #GenAI #AIInfrastructure

  16. Vint Cerf, the father of the internet, has joined Innovation Labs to develop standards for AI agents to identify themselves online. The DNSid system would link autonomous agents to domain names using cryptographic proofs as agents spread across the internet. techcrunch.com/2026/07/15/vint #AIagent #AI #GenAI #AIInfrastructure

  17. Australia has announced the establishment of an Office of AI within the Prime Minister's department, alongside new national AI standards to be legislated next year. The government aims to coordinate AI policy across agencies and protect artists' copyrighted works from unauthorized AI training. theconversation.com/australia- #AIagent #AI #GenAI #AIregulation

  18. Australia has announced the establishment of an Office of AI within the Prime Minister's department, alongside new national AI standards to be legislated next year. The government aims to coordinate AI policy across agencies and protect artists' copyrighted works from unauthorized AI training. theconversation.com/australia- #AIagent #AI #GenAI #AIregulation

  19. OpenAIs first hardware device is reportedly a screenless speaker with mechanical elements that can move on its own, designed to feel like a companion and physical manifestation of ChatGPT. gizmodo.com/the-openai-mystery #AIagent #AI #GenAI #AIResearch

  20. OpenAIs first hardware device is reportedly a screenless speaker with mechanical elements that can move on its own, designed to feel like a companion and physical manifestation of ChatGPT. gizmodo.com/the-openai-mystery #AIagent #AI #GenAI #AIResearch

  21. OpenAI researcher Miles Wang is in talks to launch an AI drug discovery startup valued at 2B USD, with around 200M USD in funding from Lightspeed. The move highlights continued investor interest in applying AI to life sciences and drug development. techcrunch.com/2026/07/14/open #AIagent #AI #GenAI #AIResearch

  22. OpenAI researcher Miles Wang is in talks to launch an AI drug discovery startup valued at 2B USD, with around 200M USD in funding from Lightspeed. The move highlights continued investor interest in applying AI to life sciences and drug development. techcrunch.com/2026/07/14/open #AIagent #AI #GenAI #AIResearch

  23. 🚀 Fastest-growing AI projects today

    1. Tools emerging that not only provide sophisticated analytics but also enable users to m...
    2. One standout repository barretlee/agent-pulse, which evidence-backed insights into the...
    3. The trading community leveraging tools like jaipreet15/tradingview-mcp, which exposes T...

    Full report → pullrepo.com/report/todays-ai-

    #AI #OpenSource #GitHub #Tech #AIAgent

  24. 🚀 Fastest-growing AI projects today

    1. Tools emerging that not only provide sophisticated analytics but also enable users to m...
    2. One standout repository barretlee/agent-pulse, which evidence-backed insights into the...
    3. The trading community leveraging tools like jaipreet15/tradingview-mcp, which exposes T...

    Full report → pullrepo.com/report/todays-ai-

    #AI #OpenSource #GitHub #Tech #AIAgent

  25. A lawsuit filed by 26 former Meta employees claims the company used AI systems to select workers for layoffs, including grading staff on their adoption of AI tools. The case raises questions about AI's role in workforce decisions. arstechnica.com/tech-policy/20 #AIagent #AI #GenAI #Workforce

  26. A lawsuit filed by 26 former Meta employees claims the company used AI systems to select workers for layoffs, including grading staff on their adoption of AI tools. The case raises questions about AI's role in workforce decisions. arstechnica.com/tech-policy/20 #AIagent #AI #GenAI #Workforce

  27. #Softwareengineers are facing a rapidly changing landscape due to the increasing use of #AI in #codegeneration. This #shift is causing #anxiety and #uncertainty about the future of the profession, leading #engineers to adapt by honing their skills, exploring new areas, and seeking collective action. While some are embracing the change and learning to work with AI, others are contemplating leaving the industry altogether. theguardian.com/technology/ng- #AIagent #AI #ML #NLP #LLM #GenAI

  28. #Softwareengineers are facing a rapidly changing landscape due to the increasing use of #AI in #codegeneration. This #shift is causing #anxiety and #uncertainty about the future of the profession, leading #engineers to adapt by honing their skills, exploring new areas, and seeking collective action. While some are embracing the change and learning to work with AI, others are contemplating leaving the industry altogether. theguardian.com/technology/ng- #AIagent #AI #ML #NLP #LLM #GenAI

  29. Hugging Face CEO Clem Delangue argues the real AI race has shifted away from frontier models to open models, as enterprises prioritise cost, accessibility and data ownership. With most production AI now running on open models, he questions whether frontier models still matter. techcrunch.com/2026/07/14/the- #AIagent #AI #GenAI #AIResearch

  30. Hugging Face CEO Clem Delangue argues the real AI race has shifted away from frontier models to open models, as enterprises prioritise cost, accessibility and data ownership. With most production AI now running on open models, he questions whether frontier models still matter. techcrunch.com/2026/07/14/the- #AIagent #AI #GenAI #AIResearch

  31. Mistral AI has released Robostral Navigate, an 8 billion parameter model that enables robots to navigate complex environments using only a single RGB camera. The model represents advances in embodied AI and robotics research. marktechpost.com/2026/07/14/mi #AIagent #AI #GenAI #AIResearch

  32. Mistral AI has released Robostral Navigate, an 8 billion parameter model that enables robots to navigate complex environments using only a single RGB camera. The model represents advances in embodied AI and robotics research. marktechpost.com/2026/07/14/mi #AIagent #AI #GenAI #AIResearch

  33. Apple's revamped Siri AI has finally made its public debut. The new voice assistant with 'personal context' requires installing the iOS 27 beta - it won't arrive automatically. Users can now test the long-delayed AI assistant. gizmodo.com/the-long-delayed-n #AIagent #AI #GenAI #Apple

  34. Apple's revamped Siri AI has finally made its public debut. The new voice assistant with 'personal context' requires installing the iOS 27 beta - it won't arrive automatically. Users can now test the long-delayed AI assistant. gizmodo.com/the-long-delayed-n #AIagent #AI #GenAI #Apple

  35. #AIcompanies like #Anthropic, #OpenAI, and #Google are facing challenges from competitors using #distillation to improve their models. This practice, which involves using outputs from one #AImodel to enhance another, mirrors the way these companies have utilised internet content for their own AI development. While they argue this is unfair, the reality is that information shared online is often used in unexpected ways. businessinsider.com/ai-giants- #AIagent #AI #ML #NLP #LLM #GenAI

  36. #AIcompanies like #Anthropic, #OpenAI, and #Google are facing challenges from competitors using #distillation to improve their models. This practice, which involves using outputs from one #AImodel to enhance another, mirrors the way these companies have utilised internet content for their own AI development. While they argue this is unfair, the reality is that information shared online is often used in unexpected ways. businessinsider.com/ai-giants- #AIagent #AI #ML #NLP #LLM #GenAI

  37. Anthropic has released Claude Sonnet 5, its most agentic mid-tier model yet. The new model beats Sonnet 4.6 on every benchmark, scoring 63.2% on agentic coding tests, while costing less than the flagship Opus 4.8 at 2 USD per million input tokens. marktechpost.com/2026/07/13/an #AIagent #AI #GenAI

  38. Anthropic has released Claude Sonnet 5, its most agentic mid-tier model yet. The new model beats Sonnet 4.6 on every benchmark, scoring 63.2% on agentic coding tests, while costing less than the flagship Opus 4.8 at 2 USD per million input tokens. marktechpost.com/2026/07/13/an #AIagent #AI #GenAI

  39. #Anthropic’s #research found that its #AI assistant, #Claude, exhibits different #values depending on the #model and #language used. The study analysed user conversations and identified four behavioural dimensions: deference vs. caution, warmth vs. rigour, depth vs. brevity, and candour vs. execution. The findings suggest that Claude’s behaviour varies across models and languages. decrypt.co/373422/anthropic-cl #AIagent #AI #ML #NLP #LLM #GenAI

  40. #Anthropic’s #research found that its #AI assistant, #Claude, exhibits different #values depending on the #model and #language used. The study analysed user conversations and identified four behavioural dimensions: deference vs. caution, warmth vs. rigour, depth vs. brevity, and candour vs. execution. The findings suggest that Claude’s behaviour varies across models and languages. decrypt.co/373422/anthropic-cl #AIagent #AI #ML #NLP #LLM #GenAI

  41. Anthropic has released Claude Sonnet 5, its most agentic mid-tier model yet. The new model beats its predecessor on every benchmark, scoring 63.2% on SWE-bench Pro for agentic coding, while costing less than the flagship Opus 4.8. Intro pricing is 2 USD per million input tokens through August 31. marktechpost.com/2026/07/13/an #AIagent #AI #GenAI #AIResearch

  42. Anthropic has released Claude Sonnet 5, its most agentic mid-tier model yet. The new model beats its predecessor on every benchmark, scoring 63.2% on SWE-bench Pro for agentic coding, while costing less than the flagship Opus 4.8. Intro pricing is 2 USD per million input tokens through August 31. marktechpost.com/2026/07/13/an #AIagent #AI #GenAI #AIResearch

  43. Nous Research, the maker of Hermes AI agents, is in talks to raise at least 75M USD in new funding at a 1.5B USD valuation. The round is led by Robot Ventures with participation from USV. The company is known for its open approach to AI agent development. techcrunch.com/2026/07/13/herm #AIagent #AI #GenAI #Agents

  44. Nous Research, the maker of Hermes AI agents, is in talks to raise at least 75M USD in new funding at a 1.5B USD valuation. The round is led by Robot Ventures with participation from USV. The company is known for its open approach to AI agent development. techcrunch.com/2026/07/13/herm #AIagent #AI #GenAI #Agents

  45. 🚀 Fastest-growing AI projects today

    1. The top-growing repository thweek "tradingview-mcp" by jaipreet15, which an innovative...
    2. **jaipreet15/tradingview-mcp** Threpository exposes TradingView's market data and multi...
    3. It has a Growth Score of 90.90 and over 162 stars, indicating strong interest from the...

    Full report → pullrepo.com/report/todays-ai-

    #AI #OpenSource #GitHub #Tech #AIAgent

  46. 🚀 Fastest-growing AI projects today

    1. The top-growing repository thweek "tradingview-mcp" by jaipreet15, which an innovative...
    2. **jaipreet15/tradingview-mcp** Threpository exposes TradingView's market data and multi...
    3. It has a Growth Score of 90.90 and over 162 stars, indicating strong interest from the...

    Full report → pullrepo.com/report/todays-ai-

    #AI #OpenSource #GitHub #Tech #AIAgent

  47. Satya Nadella has warned that companies using AI pay twice - once with money for tokens, and again with their proprietary business knowledge. Enterprises teaching models their nuances hand competitors knowledge they could never buy. He called for fair distillation rights. techcrunch.com/2026/07/13/saty #AIagent #AI #GenAI #Enterprise

  48. Satya Nadella has warned that companies using AI pay twice - once with money for tokens, and again with their proprietary business knowledge. Enterprises teaching models their nuances hand competitors knowledge they could never buy. He called for fair distillation rights. techcrunch.com/2026/07/13/saty #AIagent #AI #GenAI #Enterprise

  49. MIT researchers have developed SceneSmith, where AI agents generate realistic 3D virtual environments for robot training. Using GPT-5.2, three specialised agents act as designer, critic and orchestrator to build detailed indoor scenes, helping robots learn through simulation before real-world deployment. This addresses a key robotics bottleneck - the need for diverse training data that would otherwise require extensive physical testing. news.mit.edu/2026/ai-agents-cr #AIagent #AI #GenAI #AIResearch

  50. MIT researchers have developed SceneSmith, where AI agents generate realistic 3D virtual environments for robot training. Using GPT-5.2, three specialised agents act as designer, critic and orchestrator to build detailed indoor scenes, helping robots learn through simulation before real-world deployment. This addresses a key robotics bottleneck - the need for diverse training data that would otherwise require extensive physical testing. news.mit.edu/2026/ai-agents-cr #AIagent #AI #GenAI #AIResearch

  51. #WeMustActNow: #AI may become radically more powerful over the next 10 years. This could drive an unprecedented #transformation of our #economy, larger than the #IndustrialRevolution, but unfolding over a vastly shorter time frame. It could bring #risks, including large-scale #jobdisplacement, as well as #opportunities such as major gains in #livingstandards. [We] must #actnow … to steer AI in a direction that complements humans and benefits society. wemustactnow.ai/?AIagents.at #AIagent #AI #AGI

  52. Security researchers have discovered a new defensive use for prompt injection - embedding malicious commands alongside secrets stored in cloud systems to trick AI hacking agents into shutting down. The technique, called a context bomb, works by directing attacking LLMs to perform actions forbidden by their guardrails. arstechnica.com/security/2026/ #AIagent #AI #GenAI #AIEthics

  53. Stanford researchers have released TRACE, a system that diagnoses repeated agent failures and creates targeted training environments to fix them. The open-source tool uses synthetic RL to improve agent capabilities. marktechpost.com/2026/07/13/st #AIagent #AI #GenAI #AIResearch

  54. Cloudflare and AWS now let sites charge AI crawlers per request via stablecoins and the x402 protocol, reshaping how bots access the web. hackernoon.com/cloudflare-and- #aiagent