The AI Track
@theaitrack.bsky.social
AI Coverage: Tracking what’s next in AI and how will it impact us.
From the editors of https://theaitrack.com/
From the editors of https://theaitrack.com/
A Nature study (Oct 2025) by UC Berkeley Haas, Stanford, and Oxford finds women are systematically portrayed as younger than men across 1.4M online images and AI models. Algorithms like ChatGPT amplify the bias, reshaping hiring and perceptions.
🔗 Source: tinyurl.com/bp6renbs
#AI #Bias #Research
🔗 Source: tinyurl.com/bp6renbs
#AI #Bias #Research
October 12, 2025 at 7:46 AM
A Nature study (Oct 2025) by UC Berkeley Haas, Stanford, and Oxford finds women are systematically portrayed as younger than men across 1.4M online images and AI models. Algorithms like ChatGPT amplify the bias, reshaping hiring and perceptions.
🔗 Source: tinyurl.com/bp6renbs
#AI #Bias #Research
🔗 Source: tinyurl.com/bp6renbs
#AI #Bias #Research
OpenAI releases a political-bias evaluation for ChatGPT: ~500 prompts across 100 topics, scored on 5 axes. GPT-5 reduces bias by ~30% vs prior models. In production analysis, <0.01% of responses show signs of political bias. Web search excluded.
🔗 More: tinyurl.com/bdfvstkd
#AI #OpenAI #ChatGPT
🔗 More: tinyurl.com/bdfvstkd
#AI #OpenAI #ChatGPT
October 11, 2025 at 2:12 PM
OpenAI releases a political-bias evaluation for ChatGPT: ~500 prompts across 100 topics, scored on 5 axes. GPT-5 reduces bias by ~30% vs prior models. In production analysis, <0.01% of responses show signs of political bias. Web search excluded.
🔗 More: tinyurl.com/bdfvstkd
#AI #OpenAI #ChatGPT
🔗 More: tinyurl.com/bdfvstkd
#AI #OpenAI #ChatGPT
Oxford, Google Cloud, and Radboud University show Gemini AI can classify cosmic events with ~93% accuracy using just 15 examples. Published in Nature Astronomy, the study highlights transparent AI as a new tool for astronomy.
#AI #Astronomy #Oxford
🔗 Study: www.nature.com/articles/s41...
#AI #Astronomy #Oxford
🔗 Study: www.nature.com/articles/s41...
October 11, 2025 at 8:52 AM
Oxford, Google Cloud, and Radboud University show Gemini AI can classify cosmic events with ~93% accuracy using just 15 examples. Published in Nature Astronomy, the study highlights transparent AI as a new tool for astronomy.
#AI #Astronomy #Oxford
🔗 Study: www.nature.com/articles/s41...
#AI #Astronomy #Oxford
🔗 Study: www.nature.com/articles/s41...
OpenAI/Harvard/Duke study: 700M weekly ChatGPT users (10% of adults). 73% of queries are non-work. Writing = 42% of work use. Under-26s send 46% of all messages. Growth fastest in $10k–40k GDP countries. #ChatGPT #AI #TechNews
🔗 Report: tinyurl.com/hrpuavu9
🔗 Report: tinyurl.com/hrpuavu9
October 10, 2025 at 10:17 PM
OpenAI/Harvard/Duke study: 700M weekly ChatGPT users (10% of adults). 73% of queries are non-work. Writing = 42% of work use. Under-26s send 46% of all messages. Growth fastest in $10k–40k GDP countries. #ChatGPT #AI #TechNews
🔗 Report: tinyurl.com/hrpuavu9
🔗 Report: tinyurl.com/hrpuavu9
🌐🚀 This Week’s Top 7 AI Updates You Need to Know
🔗 Read all the important key information on The AI Track: tinyurl.com/3a3bks2s
👍 Follow us for continuous insights and updates in the world of AI.
#AIWeeklyNews #AI #artificialintelligence #TechTrends #TheAITrack #aibusiness #ainews #technews
🔗 Read all the important key information on The AI Track: tinyurl.com/3a3bks2s
👍 Follow us for continuous insights and updates in the world of AI.
#AIWeeklyNews #AI #artificialintelligence #TechTrends #TheAITrack #aibusiness #ainews #technews
October 6, 2025 at 7:25 AM
🌐🚀 This Week’s Top 7 AI Updates You Need to Know
🔗 Read all the important key information on The AI Track: tinyurl.com/3a3bks2s
👍 Follow us for continuous insights and updates in the world of AI.
#AIWeeklyNews #AI #artificialintelligence #TechTrends #TheAITrack #aibusiness #ainews #technews
🔗 Read all the important key information on The AI Track: tinyurl.com/3a3bks2s
👍 Follow us for continuous insights and updates in the world of AI.
#AIWeeklyNews #AI #artificialintelligence #TechTrends #TheAITrack #aibusiness #ainews #technews
🌐🚀 This Week’s Top 9 AI Updates You Need to Know: 📰💡
1️⃣ 🏆 Gemini 2.5 Deep Think Wins Gold at ICPC World Finals
Google’s Gemini 2.5 Deep Think model solved 10 of 12 problems in Baku, even outperforming all human teams on one task — a milestone DeepMind called “historic.”
1️⃣ 🏆 Gemini 2.5 Deep Think Wins Gold at ICPC World Finals
Google’s Gemini 2.5 Deep Think model solved 10 of 12 problems in Baku, even outperforming all human teams on one task — a milestone DeepMind called “historic.”
September 22, 2025 at 10:57 AM
🌐🚀 This Week’s Top 9 AI Updates You Need to Know: 📰💡
1️⃣ 🏆 Gemini 2.5 Deep Think Wins Gold at ICPC World Finals
Google’s Gemini 2.5 Deep Think model solved 10 of 12 problems in Baku, even outperforming all human teams on one task — a milestone DeepMind called “historic.”
1️⃣ 🏆 Gemini 2.5 Deep Think Wins Gold at ICPC World Finals
Google’s Gemini 2.5 Deep Think model solved 10 of 12 problems in Baku, even outperforming all human teams on one task — a milestone DeepMind called “historic.”
🩺 An AI-powered stethoscope, developed by Imperial College London and Eko Health, can quickly detect three major heart conditions.
🩺 It performs both an ECG (electrocardiogram) and uses a microphone to record blood flow sounds.
#AI, #AIinHealthcare
🩺 It performs both an ECG (electrocardiogram) and uses a microphone to record blood flow sounds.
#AI, #AIinHealthcare
September 5, 2025 at 2:48 PM
🩺 An AI-powered stethoscope, developed by Imperial College London and Eko Health, can quickly detect three major heart conditions.
🩺 It performs both an ECG (electrocardiogram) and uses a microphone to record blood flow sounds.
#AI, #AIinHealthcare
🩺 It performs both an ECG (electrocardiogram) and uses a microphone to record blood flow sounds.
#AI, #AIinHealthcare
🌐🚀 This Week’s Top 9 AI Headlines 📰💡
🔗 Full details on The AI Track 🔗 theaitrack.com/ai-news-augu...
👍 Follow us for continuous insights and updates in the world of AI.
🔗 Full details on The AI Track 🔗 theaitrack.com/ai-news-augu...
👍 Follow us for continuous insights and updates in the world of AI.
August 10, 2025 at 6:05 PM
🌐🚀 This Week’s Top 9 AI Headlines 📰💡
🔗 Full details on The AI Track 🔗 theaitrack.com/ai-news-augu...
👍 Follow us for continuous insights and updates in the world of AI.
🔗 Full details on The AI Track 🔗 theaitrack.com/ai-news-augu...
👍 Follow us for continuous insights and updates in the world of AI.
🌶 Elon Musk’s xAI released Grok-Imagine, an AI tool that generates images and short videos including NSFW content via a “Spicy” mode.
🌶 The tool supports text-to-image and image-to-video conversion with voice input, raising many concerns. #AI
🌶 More: theaitrack.com/xai-grok-ima...
🌶 The tool supports text-to-image and image-to-video conversion with voice input, raising many concerns. #AI
🌶 More: theaitrack.com/xai-grok-ima...
August 8, 2025 at 7:17 PM
🌶 Elon Musk’s xAI released Grok-Imagine, an AI tool that generates images and short videos including NSFW content via a “Spicy” mode.
🌶 The tool supports text-to-image and image-to-video conversion with voice input, raising many concerns. #AI
🌶 More: theaitrack.com/xai-grok-ima...
🌶 The tool supports text-to-image and image-to-video conversion with voice input, raising many concerns. #AI
🌶 More: theaitrack.com/xai-grok-ima...
🌐🚀 This Week’s Top 5 AI Updates You Need to Know 📰💡
🔗 Read all the important key information on The AI Track 🔗 theaitrack.com/ai-news-augu...
🔗 Read all the important key information on The AI Track 🔗 theaitrack.com/ai-news-augu...
August 3, 2025 at 8:45 AM
🌐🚀 This Week’s Top 5 AI Updates You Need to Know 📰💡
🔗 Read all the important key information on The AI Track 🔗 theaitrack.com/ai-news-augu...
🔗 Read all the important key information on The AI Track 🔗 theaitrack.com/ai-news-augu...
🌐🚀 This Week’s Top 9 AI Updates You Need to Know: 📰💡
1️⃣ 🧠 OpenAI Releases ChatGPT Agent With AI Task Execution (“Agent Mode”)
2️⃣ 🔞 xAI Adds NSFW Anime “Companion” to Grok, Accessible Even in Kid Mode
3️⃣ 🔐 Google's "Big Sleep" AI Agent Foils In-the-Wild Vulnerability Exploitation
#AI
1️⃣ 🧠 OpenAI Releases ChatGPT Agent With AI Task Execution (“Agent Mode”)
2️⃣ 🔞 xAI Adds NSFW Anime “Companion” to Grok, Accessible Even in Kid Mode
3️⃣ 🔐 Google's "Big Sleep" AI Agent Foils In-the-Wild Vulnerability Exploitation
#AI
July 20, 2025 at 9:02 PM
🌐🚀 This Week’s Top 9 AI Updates You Need to Know: 📰💡
1️⃣ 🧠 OpenAI Releases ChatGPT Agent With AI Task Execution (“Agent Mode”)
2️⃣ 🔞 xAI Adds NSFW Anime “Companion” to Grok, Accessible Even in Kid Mode
3️⃣ 🔐 Google's "Big Sleep" AI Agent Foils In-the-Wild Vulnerability Exploitation
#AI
1️⃣ 🧠 OpenAI Releases ChatGPT Agent With AI Task Execution (“Agent Mode”)
2️⃣ 🔞 xAI Adds NSFW Anime “Companion” to Grok, Accessible Even in Kid Mode
3️⃣ 🔐 Google's "Big Sleep" AI Agent Foils In-the-Wild Vulnerability Exploitation
#AI
🏎️💻 Data scientist Mariana Antaya's AI model has accurately predicted 3 out of 5 F1 race winners this season, including Norris in Australia and Verstappen in Japan. The model's precision is improving with each race! #F1 #AI #MachineLearning #Formula1
May 20, 2025 at 7:02 PM
🏎️💻 Data scientist Mariana Antaya's AI model has accurately predicted 3 out of 5 F1 race winners this season, including Norris in Australia and Verstappen in Japan. The model's precision is improving with each race! #F1 #AI #MachineLearning #Formula1
🧠 New BCI tech allows people with paralysis to "custom-design" how touch feels — from "smooth keys" to "furry cats". Study showed 35% object ID success using only sensation. A real leap for neuroprosthetics.
#Neurotech #BCI #ArtificialTouch #Paralysis #ScienceNews
#Neurotech #BCI #ArtificialTouch #Paralysis #ScienceNews
May 13, 2025 at 5:53 AM
🧠 New BCI tech allows people with paralysis to "custom-design" how touch feels — from "smooth keys" to "furry cats". Study showed 35% object ID success using only sensation. A real leap for neuroprosthetics.
#Neurotech #BCI #ArtificialTouch #Paralysis #ScienceNews
#Neurotech #BCI #ArtificialTouch #Paralysis #ScienceNews
🧬 Harvard & MIT introduce TxAgent: an AI agent leveraging 211 biomedical tools to deliver personalized treatment recommendations with 92.1% accuracy. #AI #HealthcareInnovation #TxAgent
May 12, 2025 at 8:09 PM
🧬 Harvard & MIT introduce TxAgent: an AI agent leveraging 211 biomedical tools to deliver personalized treatment recommendations with 92.1% accuracy. #AI #HealthcareInnovation #TxAgent
🚀 The Top 10 Most Wanted AI Skills in 2025
1️⃣ Python (199K postings)
2️⃣ Data science (833% growth!)
3️⃣ AWS (1,778% surge!)
Soft skills like Agile & project management also made the cut!
📈 Job postings up 20% in 2024.
📖 Source: 2025 AI Index Report www.visualcapitalist.com/the-top-10-m...
#AI
1️⃣ Python (199K postings)
2️⃣ Data science (833% growth!)
3️⃣ AWS (1,778% surge!)
Soft skills like Agile & project management also made the cut!
📈 Job postings up 20% in 2024.
📖 Source: 2025 AI Index Report www.visualcapitalist.com/the-top-10-m...
#AI
April 30, 2025 at 6:11 AM
🚀 The Top 10 Most Wanted AI Skills in 2025
1️⃣ Python (199K postings)
2️⃣ Data science (833% growth!)
3️⃣ AWS (1,778% surge!)
Soft skills like Agile & project management also made the cut!
📈 Job postings up 20% in 2024.
📖 Source: 2025 AI Index Report www.visualcapitalist.com/the-top-10-m...
#AI
1️⃣ Python (199K postings)
2️⃣ Data science (833% growth!)
3️⃣ AWS (1,778% surge!)
Soft skills like Agile & project management also made the cut!
📈 Job postings up 20% in 2024.
📖 Source: 2025 AI Index Report www.visualcapitalist.com/the-top-10-m...
#AI
🎙️ Sydney’s CADA radio used an AI-generated host for 6 months without telling listeners. Created with ElevenLabs, “Thy” ran Mon–Fri, 11–3, before fans discovered she wasn’t real. No rules broken—but trust was. #AI #Radio #MediaEthics #ElevenLabs #Australia #SyntheticVoices
April 29, 2025 at 6:25 PM
🎙️ Sydney’s CADA radio used an AI-generated host for 6 months without telling listeners. Created with ElevenLabs, “Thy” ran Mon–Fri, 11–3, before fans discovered she wasn’t real. No rules broken—but trust was. #AI #Radio #MediaEthics #ElevenLabs #Australia #SyntheticVoices
𝗙𝗿𝗼𝗺 𝗜𝗹𝘆𝗮 𝗦𝘂𝘁𝘀𝗸𝗲𝘃𝗲𝗿 𝗮𝘁 𝗡𝗲𝘂𝗿𝗜𝗣𝗦 🌐
“Pre-training is ending—we’ve reached peak data," says Ilya Sutskever. 🤖 AI systems will be more agentic, understanding from limited data. As AI evolves, the future is unpredictable, but who knows—maybe AIs will just want to coexist with us and have rights! 🧠💭
#AI
“Pre-training is ending—we’ve reached peak data," says Ilya Sutskever. 🤖 AI systems will be more agentic, understanding from limited data. As AI evolves, the future is unpredictable, but who knows—maybe AIs will just want to coexist with us and have rights! 🧠💭
#AI
April 16, 2025 at 7:51 AM
𝗙𝗿𝗼𝗺 𝗜𝗹𝘆𝗮 𝗦𝘂𝘁𝘀𝗸𝗲𝘃𝗲𝗿 𝗮𝘁 𝗡𝗲𝘂𝗿𝗜𝗣𝗦 🌐
“Pre-training is ending—we’ve reached peak data," says Ilya Sutskever. 🤖 AI systems will be more agentic, understanding from limited data. As AI evolves, the future is unpredictable, but who knows—maybe AIs will just want to coexist with us and have rights! 🧠💭
#AI
“Pre-training is ending—we’ve reached peak data," says Ilya Sutskever. 🤖 AI systems will be more agentic, understanding from limited data. As AI evolves, the future is unpredictable, but who knows—maybe AIs will just want to coexist with us and have rights! 🧠💭
#AI
The Children’s Hospital of Eastern Ontario (CHEO) is Canada's first hospital using AI to diagnose rare diseases, by analyzing electronic health records. The system addresses long-standing diagnostic delays by integrating AI at the beginning of care pathways. #AI
Source: tinyurl.com/33v66cr4
Source: tinyurl.com/33v66cr4
April 14, 2025 at 8:13 PM
The Children’s Hospital of Eastern Ontario (CHEO) is Canada's first hospital using AI to diagnose rare diseases, by analyzing electronic health records. The system addresses long-standing diagnostic delays by integrating AI at the beginning of care pathways. #AI
Source: tinyurl.com/33v66cr4
Source: tinyurl.com/33v66cr4
𝗛𝗲𝗮𝘁 𝗦𝘁𝗿𝗼𝗸𝗲 𝗗𝗲𝘁𝗲𝗰𝘁𝗲𝗱 𝗧𝗵𝗿𝗼𝘂𝗴𝗵 𝗧𝘄𝗲𝗲𝘁𝘀 🌞🤖
Transformer models like LUKE Japanese base lite are now classifying heat stroke tweets in Japan with 85.52% accuracy 🌡️ This method offers real-time public health surveillance - faster responses to heat strokes & other crises.
Source: tinyurl.com/yz6h4ynv
#AI
Transformer models like LUKE Japanese base lite are now classifying heat stroke tweets in Japan with 85.52% accuracy 🌡️ This method offers real-time public health surveillance - faster responses to heat strokes & other crises.
Source: tinyurl.com/yz6h4ynv
#AI
April 14, 2025 at 7:06 AM
𝗛𝗲𝗮𝘁 𝗦𝘁𝗿𝗼𝗸𝗲 𝗗𝗲𝘁𝗲𝗰𝘁𝗲𝗱 𝗧𝗵𝗿𝗼𝘂𝗴𝗵 𝗧𝘄𝗲𝗲𝘁𝘀 🌞🤖
Transformer models like LUKE Japanese base lite are now classifying heat stroke tweets in Japan with 85.52% accuracy 🌡️ This method offers real-time public health surveillance - faster responses to heat strokes & other crises.
Source: tinyurl.com/yz6h4ynv
#AI
Transformer models like LUKE Japanese base lite are now classifying heat stroke tweets in Japan with 85.52% accuracy 🌡️ This method offers real-time public health surveillance - faster responses to heat strokes & other crises.
Source: tinyurl.com/yz6h4ynv
#AI
AI Detects Sepsis 6 Hours Earlier 🚑🤖
An AI system developed at Johns Hopkins Univ. Tested across 590,000 patients, it boasts 40% accuracy in detecting 82% of sepsis cases, potentially saving thousands of lives annually. Major leap from past methods. 🕒
Source: tinyurl.com/y89znwsj
#AI
An AI system developed at Johns Hopkins Univ. Tested across 590,000 patients, it boasts 40% accuracy in detecting 82% of sepsis cases, potentially saving thousands of lives annually. Major leap from past methods. 🕒
Source: tinyurl.com/y89znwsj
#AI
April 14, 2025 at 7:02 AM
AI Detects Sepsis 6 Hours Earlier 🚑🤖
An AI system developed at Johns Hopkins Univ. Tested across 590,000 patients, it boasts 40% accuracy in detecting 82% of sepsis cases, potentially saving thousands of lives annually. Major leap from past methods. 🕒
Source: tinyurl.com/y89znwsj
#AI
An AI system developed at Johns Hopkins Univ. Tested across 590,000 patients, it boasts 40% accuracy in detecting 82% of sepsis cases, potentially saving thousands of lives annually. Major leap from past methods. 🕒
Source: tinyurl.com/y89znwsj
#AI
April 12, 2025 at 8:34 AM
Abductive Reasoning: The Creative and Deceptive Brain of AI 🤖💭
AI systems like OpenAI’s o1 demonstrate creativity & complex reasoning, but their underlying processes are mysterious—even to their creators. 🤯 #AI #ArtificialIntelligence
More: tinyurl.com/53bxkssy
AI systems like OpenAI’s o1 demonstrate creativity & complex reasoning, but their underlying processes are mysterious—even to their creators. 🤯 #AI #ArtificialIntelligence
More: tinyurl.com/53bxkssy
April 12, 2025 at 8:30 AM
Abductive Reasoning: The Creative and Deceptive Brain of AI 🤖💭
AI systems like OpenAI’s o1 demonstrate creativity & complex reasoning, but their underlying processes are mysterious—even to their creators. 🤯 #AI #ArtificialIntelligence
More: tinyurl.com/53bxkssy
AI systems like OpenAI’s o1 demonstrate creativity & complex reasoning, but their underlying processes are mysterious—even to their creators. 🤯 #AI #ArtificialIntelligence
More: tinyurl.com/53bxkssy
AI-Designed Proteins to Combat Snake Venom 🐍🧬
AI is now designing proteins that neutralize venom (found in snakes like cobras, taipans, and mambas), offering hope for more accessible & cost-effective treatments for snakebites that claims over 100,000 lives annually💡
Source: tinyurl.com/mvpbucpp
#AI
AI is now designing proteins that neutralize venom (found in snakes like cobras, taipans, and mambas), offering hope for more accessible & cost-effective treatments for snakebites that claims over 100,000 lives annually💡
Source: tinyurl.com/mvpbucpp
#AI
April 11, 2025 at 9:21 AM
AI-Designed Proteins to Combat Snake Venom 🐍🧬
AI is now designing proteins that neutralize venom (found in snakes like cobras, taipans, and mambas), offering hope for more accessible & cost-effective treatments for snakebites that claims over 100,000 lives annually💡
Source: tinyurl.com/mvpbucpp
#AI
AI is now designing proteins that neutralize venom (found in snakes like cobras, taipans, and mambas), offering hope for more accessible & cost-effective treatments for snakebites that claims over 100,000 lives annually💡
Source: tinyurl.com/mvpbucpp
#AI
🧠🔬AI Enhances Breast Cancer Detection by 17.6% 💡💖
🧠🔬 In a massive study (460K+ women), AI-assisted mammograms detected 17.6% more breast cancer cases, reduced false positives, and flagged suspicious cases, all while aiding radiologists.🧠🔬 #AIHealth
Source: tinyurl.com/2fvnx4vj
🧠🔬 In a massive study (460K+ women), AI-assisted mammograms detected 17.6% more breast cancer cases, reduced false positives, and flagged suspicious cases, all while aiding radiologists.🧠🔬 #AIHealth
Source: tinyurl.com/2fvnx4vj
April 11, 2025 at 9:19 AM
🧠🔬AI Enhances Breast Cancer Detection by 17.6% 💡💖
🧠🔬 In a massive study (460K+ women), AI-assisted mammograms detected 17.6% more breast cancer cases, reduced false positives, and flagged suspicious cases, all while aiding radiologists.🧠🔬 #AIHealth
Source: tinyurl.com/2fvnx4vj
🧠🔬 In a massive study (460K+ women), AI-assisted mammograms detected 17.6% more breast cancer cases, reduced false positives, and flagged suspicious cases, all while aiding radiologists.🧠🔬 #AIHealth
Source: tinyurl.com/2fvnx4vj
🎶 The fusion of AI and music is taking a bold step forward with AI-driven voice de-aging. Artists can now rejuvenate their vocals, bringing back the tones of their youth. #AI
April 4, 2025 at 7:08 AM
🎶 The fusion of AI and music is taking a bold step forward with AI-driven voice de-aging. Artists can now rejuvenate their vocals, bringing back the tones of their youth. #AI