The hottest AI Research Substack posts right now

And their main takeaways
Category
Top Technology Topics

The Sequence Radar #775: Last Week in AI: Tokens, Throughput, and Trillions

TheSequence • 63 implied HN points • 21 Dec 25
🕹 Technology AI Research
  1. Massive funding and infrastructure bets are setting the rules: the companies that can industrialize models into cheap, reliable global services will win more than those with just the fanciest research demos.
  2. Engineering focus has shifted to throughput, latency, and long-context agentic capabilities, with new models and hardware optimized to move lots of tokens through multi-step workflows at predictable cost.
  3. Generative outputs and developer workflows are becoming iterative and productized — image editing in chat and tightened data/observability loops make AI a usable creative IDE, while enterprise platforms race to own the data plane and production tooling.

Import AI 359: $1 billion gov supercomputer; Apple’s good synthetic data technique; and a thousand-year old data library

Import AI • 339 implied HN points • 05 Feb 24
🕹 Technology AI Research
  1. Google uses LLM-powered bug fixing that is more efficient than human fixes, highlighting the impact of AI integration in speeding up processes.
  2. Yoshua Bengio suggests governments invest in supercomputers for AI development to stay ahead in monitoring tech giants, emphasizing the importance of AI investment in the public sector.
  3. Microsoft's Project Silica showcases a long-term storage solution using glass for archiving data, which is a unique and durable alternative to traditional methods.
  4. Apple's WRAP technique creates synthetic data effectively by rephrasing web articles, enhancing model performance and showcasing the value of incorporating synthetic data in training.

Import AI 349: Distributed training breaks AI policy; turning GPT4 bad for $245; better weather forecasting through AI

Import AI • 459 implied HN points • 20 Nov 23
🕹 Technology AI Research
  1. Graph Neural Networks are used to create an advanced weather forecasting system called GraphCast, outperforming traditional weather simulation.
  2. Open Philanthropy offers grants to evaluate large language models like LLM agents for real-world tasks, exploring potential safety risks and impacts.
  3. Neural MMO 2.0 platform enables training AI agents in complex multiplayer games, showcasing the evolving landscape of AI research beyond language models.

Towards American Truly Open Models: The ATOM Project

Democratizing Automation • 237 implied HN points • 04 Aug 25
🕹 Technology AI Research
  1. The U.S. needs to focus on developing open AI models to regain its global leadership. This means investing in resources and creating an ecosystem that supports collaboration and research.
  2. China has been gaining ground in AI by using open models that are accessible and flexible. If the U.S. doesn't prioritize open models, American researchers and companies will look elsewhere for innovation.
  3. Building a strong network of multiple labs in the U.S. focused on open model development is crucial. This approach will help encourage growth, innovation, and diversity in AI research.

Import AI 342: Mistral dumps an LLM on BitTorrent; AMD vs NVIDIA; Sutton joins keen

Import AI • 539 implied HN points • 02 Oct 23
🕹 Technology AI Research
  1. AI startup Lamini is offering an 'LLM superstation' using AMD GPUs, challenging NVIDIA's dominance in AI chip market.
  2. AI researcher Rich Sutton has joined Keen Technologies, indicating a strong focus on developing Artificial General Intelligence (AGI).
  3. French startup Mistral released Mistral 7B, a high-quality open-source language model that outperforms other models, sparking discussions on safety measures in AI models.
Get a weekly roundup of the best Substack posts, by hacker news affinity:

Import AI 341: Neural nets can smell; technofeudalism via AI; China releases another solid open access model

Import AI • 459 implied HN points • 25 Sep 23
🕹 Technology AI Research
  1. China released open access language models trained on both English and Chinese data, emphasizing safety practices tailored to China's social context.
  2. Google and collaborators created a digital map of smells, pushing AI capabilities to not just recognize visual and audio data but also scents, opening new possibilities for exploration and understanding.
  3. An economist outlines possible societal impacts of AI advancement, predicting a future where superintelligence prompts dramatic changes in governance structures, requiring adaptability from liberal democracies.

Import AI 321: Open source GPT3; giving away democracy to AGI companies; GPT-4 is a political artifact

Import AI • 599 implied HN points • 20 Mar 23
🕹 Technology AI Research
  1. AI startup Assembly AI developed Conformer-1 using scaling laws for speech recognition domain, achieving better performance than other models.
  2. The announcement of GPT-4 by OpenAI signifies a shift towards a new political era in AI, raising concerns on the power wielded by private sector companies over AGI development.
  3. James Phillips highlights concerns over Western governments relinquishing control of AGI to US-owned private sector, proposing steps to safeguard democratic control over AI development.

Import AI 334: Better distillation; the UK's AI taskforce; money and AI

Import AI • 399 implied HN points • 10 Jul 23
🕹 Technology AI Research
  1. DeepMind developed Generalized Knowledge Distillation to make large models cheaper and more portable without losing performance.
  2. The UK's £100 million Foundation Model Taskforce aims to shape the future of safe AI and will host a global summit on AI.
  3. Significant financial investments in AI, like Databricks acquiring MosaicML for $1.3 billion, indicate growing strategic importance of AI in various sectors.

Import AI 350: Neural architecture search at Facebook scale; hunting cancer with PANDA; European VCs launch a science lab

Import AI • 279 implied HN points • 27 Nov 23
🕹 Technology AI Research
  1. An AI system called PANDA can accurately identify pancreatic cancer from scans, outperforming radiologists.
  2. Facebook developed Rankitect for neural architecture search, which has proven to create better models than human engineers alone.
  3. A European open science AI lab called Kyutai has been launched with a focus on developing large multimodal models and promoting open research.

RAG Glossary

The Tech Buffet • 179 implied HN points • 21 Jan 24
🕹 Technology AI Research
  1. Retrieval Augmented Generation (RAG) helps AI answer questions and generate content. It combines searching through documents with generating relevant answers.
  2. Using RAG can be tricky, especially in production environments. Adjustments may be needed to improve reliability and performance.
  3. Different indexing methods can optimize how RAG retrieves information. This can make it more efficient and effective in finding the right data.

Import AI 327: Stable Diffusion on phones; GPT-Hacker; UK launches a £100m AI taskforce

Import AI • 379 implied HN points • 01 May 23
🕹 Technology AI Research
  1. Google researchers optimized Stable Diffusion for efficiency on smartphones, achieving fast inference latency, a step towards industrialization of image generation.
  2. Using large language models like GPT-4 can enhance hacker capabilities, automating tasks and providing helpful tips.
  3. Political parties, like the Republican National Committee, are leveraging AI to create AI-generated content for campaigns, highlighting the emergence of AI in shaping political narratives.

How I Taught Myself to Critique AI Research [Storytime Saturdays]

Technology Made Simple • 119 implied HN points • 10 Mar 24
🕹 Technology AI Research
  1. Writing allows you to store knowledge for future reference, spot cognitive blindspots, and engage with topics more deeply for better understanding.
  2. Challenges in self-learning writing include lack of contextual understanding, a defined learning path, and a peer network for feedback.
  3. Addressing challenges in self-learning involves finding strategies to gain clarity, identifying knowledge gaps, and seeking feedback from peers.

Import AI 331: 16X smaller language models; could AMD compete with NVIDIA?; and BERT for the dark web

Import AI • 319 implied HN points • 29 May 23
🕹 Technology AI Research
  1. Researchers have found a way to significantly reduce memory requirements for training large language models, making it feasible to fine-tune on a single GPU, which could have implications for AI governance and model security.
  2. George Hotz's new company, Tiny Corp, aims to enable AMD to compete with NVIDIA in AI training chips, potentially paving the way for a more competitive AI chip market.
  3. Training language models on text from the dark web, like DarkBERT, could lead to improved detection of illicit activities online, showcasing the potential of AI systems in monitoring and identifying threats in the digital space.

Import AI 332: Mini-AI; safety through evals; Facebook releases a RLHF dataset

Import AI • 299 implied HN points • 12 Jun 23
🕹 Technology AI Research
  1. Facebook used human feedback to train its language model, BlenderBot 3x, leading to better and safer responses than its predecessor
  2. Cohere's research shows that training AI systems with specific techniques can make them easier to miniaturize, which can reduce memory requirements and latency
  3. A new organization called Apollo Research aims to develop evaluations for unsafe AI behaviors, helping improve the safety of AI companies through research into AI interpretability

Import AI 319: Sovereign AI; Facebook's weights leak on torrent networks; Google might have made a better optimizer than Adam!

Import AI • 439 implied HN points • 06 Mar 23
🕹 Technology AI Research
  1. Google researchers achieved promising results by scaling a Vision Transformer to 22B parameters, showcasing improved alignment to human visual perception.
  2. Google introduced a potentially better optimizer called Lion, showing outstanding performance across various models and tasks, including setting a new high score on ImageNet.
  3. A shift toward sovereign AI systems is emerging globally, driven by the need for countries to develop their own AI capabilities to enhance national security and economic competitiveness.

OpenAI's o1 using "search" was a PSYOP

Democratizing Automation • 435 implied HN points • 04 Dec 24
🕹 Technology AI Research
  1. OpenAI's o1 models may not actually use traditional search methods as people think. Instead, they might rely more on reinforcement learning, which is a different way of optimizing their performance.
  2. The success of OpenAI's models seems to come from using clear, measurable outcomes for training. This includes learning from mistakes and refining their approach based on feedback.
  3. OpenAI's approach focuses on scaling up the computation and training process without needing complex external search strategies. This can lead to better results by simply using the model's internal methods effectively.

The Sequence Radar #696: Google AI Ultra’s New Gold-Medal Reasoning Model is Available

TheSequence • 119 implied HN points • 03 Aug 25
🕹 Technology AI Research
  1. Google released a new AI model called Gemini 2.5 Deep Think that can solve complex math problems like a human. It performed so well that it won a gold medal at the International Math Olympiad.
  2. This model uses advanced strategies to explore many possible solutions at once, making it faster and more creative than previous AIs.
  3. The emergence of such powerful AI means we need to discuss how to use these systems responsibly, ensuring they benefit everyone and maintain fair access.

China’s DeepSeek Adds a Weird New Data Point to The AI Race

Am I Stronger Yet? • 282 implied HN points • 30 Jan 25
🕹 Technology AI Research
  1. DeepSeek's new AI model, r1, shows impressive reasoning abilities, challenging larger competitors despite its smaller budget and team. It proves that smaller companies can contribute significantly to AI advancements.
  2. The cost of training r1 was much lower than similar models, potentially signaling a shift in how AI models might be developed and run in the future. This could allow more organizations to participate in AI development without needing huge budgets.
  3. DeepSeek's approach, including releasing its model weights for public use, opens up the possibility for further research and innovation. This could change the landscape of AI by making powerful tools more accessible to everyone.

Results from poll #4

The Counterfactual • 39 implied HN points • 21 May 24
🕹 Technology AI Research
  1. The recent poll found that two topics, an explainer on interpretability and a guide to becoming an LLM-ologist, were equally popular among voters.
  2. The plan is to write about both topics in the coming months, keeping the content varied as usual.
  3. Two new papers were published this month, one on multimodal LLMs and another on Korean language models, highlighting ongoing research in these areas.

Quant Letter: January 2026, Week-3

The Parlour • 8 implied HN points • 16 Jan 26
💰 Finance AI Research
  1. Fine-tuning LLaMA-3-8B with instruction tuning and LoRA noticeably improves financial named-entity recognition, helping convert messy reports into structured data.
  2. New work on adaptive dataflow for financial time-series points to better ways to process streaming market data and boost model efficiency or accuracy.
  3. This newsletter curates recent finance ML papers and is available by subscription, with some free previews for readers who want quick research updates.

The Sequence Radar #679: From Model to Team: Several Models are Better than One: Sakana’s Blueprint for Collective AI

TheSequence • 105 implied HN points • 06 Jul 25
🕹 Technology AI Research
  1. Sakana AI has a new way to use multiple models together for better AI performance. Instead of relying on one model, they combine many to think more like humans.
  2. Their approach, called AB-MCTS, helps the AI decide whether to explore new ideas or improve current ones. This makes the AI smarter and more flexible in how it solves problems.
  3. By using several models that learn from past tasks, this system can better handle different challenges. This means AI can become more reliable and efficient in real-life applications.

Israel’s Ethnic Cleansing Push

Nonzero Newsletter • 722 implied HN points • 05 Jan 24
🌍 World Politics AI Research
  1. Concerns about Israel's possible ethnic cleansing in Gaza are getting more substantiation.
  2. AI advancements are speeding up, with predictions for various feats being revised to earlier dates.
  3. The Russia-Ukraine war is not just causing destruction, but also benefiting the military industrial complex.

Weekly Top Picks #99

The Algorithmic Bridge • 191 implied HN points • 24 Feb 25
🕹 Technology AI Research
  1. AI labs need to find the right balance between scaling their systems and efficiency in their processes.
  2. There's an AI model that criticized famous figures like Elon Musk and Donald Trump, showing it might lean towards leftist views.
  3. Tyler Cowen believes the slow integration of AI into our society is due to human limitations, not the technology itself.

Weekly Dose of Optimism #130

Not Boring by Packy McCormick • 168 implied HN points • 07 Feb 25
🕹 Technology AI Research
  1. Researchers found a new drug called CT-179 that may help stop childhood brain tumors by keeping cancer stem cells dormant. This could lead to better treatments that stop the cancer from coming back.
  2. OpenAI introduced Deep Research, a new AI that can do detailed research and create expert-level reports quickly. It's designed to help with complicated subjects, making research easier for everyone.
  3. NanoCas is a tiny CRISPR system that can edit genes in muscle and heart tissues, not just the liver. This breakthrough could help treat muscle diseases and improve gene therapies.

The Super Weight in Large Language Models

Gonzo ML • 189 implied HN points • 29 Nov 24
🕹 Technology AI Research
  1. There's a special weight in large language models called the 'super weight.' If you remove it, the model's performance crashes dramatically, showing just how crucial it is.
  2. Super weights are linked to what's called 'super activations,' meaning they help generate better text. Without them, the model struggles to create coherent sentences.
  3. Finally, researchers found ways to identify and protect these super weights during the model training and quantization processes. This makes the model more efficient and retains its quality.

Bard Becomes Gemini, Ultra is Here, and Other Ways Google is Mirroring OpenAI's Strategy

Synthedia • 59 implied HN points • 11 Feb 24
🕹 Technology AI Research
  1. Google introduced Gemini Ultra as its answer to GPT-4, integrating it into Bard to compete with ChatGPT and gain market significance.
  2. Gemini Ultra model shows strong performance in various benchmarks, outperforming GPT-4 in text, image, and reasoning tasks.
  3. Google is consolidating its AI offerings by blending Bard and Google Assistant into Gemini, aiming to provide a more advanced AI assistant experience.

DeepSeek-R1: Open model with Reasoning

Gonzo ML • 126 implied HN points • 10 Feb 25
🕹 Technology AI Research
  1. DeepSeek-R1 shows how AI models can think through problems by reasoning before giving answers. This means they can generate longer, more thoughtful responses rather than just quick answers.
  2. This model is a big step for open-source AI as it competes well with commercial versions. The community can improve it further, making powerful tools accessible for everyone.
  3. The training approach used is innovative, focusing on reinforcement learning to teach reasoning without needing a lot of examples. This could change how we train AI in the future.

Data Science Weekly - Issue 481

Data Science Weekly Newsletter • 239 implied HN points • 09 Feb 23
🕹 Technology AI Research
  1. Big Data is changing, and it's not as big a deal as we thought. Hardware is getting better faster than data sizes are growing.
  2. Research in AI can be learned just like a sport. It's about practicing skills like designing experiments and writing papers.
  3. Data Analytics can really help businesses understand their performance and make smarter decisions. It’s all about using data to solve problems and anticipate future issues.

GPT‑5.2, GLM-4.6V, Runway's GWM Worlds, GWM Avatars and GWM Robotics, Nomos 1, Devstral 2, Wan-Move, SimGym, Disco, Stripe's Agentic Commerce Suite and more

AI Brews • 10 implied HN points • 12 Dec 25
🕹 Technology AI Research
  1. Large AI models are making big leaps: new releases like GPT‑5.2 and specialized models improve reasoning, code, vision, long‑context handling, and tool use, while smaller specialist models like Nomos 1 can outperform humans on hard math tasks.
  2. Agentic and commerce-focused tools are moving into the mainstream, with products and standards that let AI agents act inside apps, make purchases, and integrate into workflows (agentic commerce, foundation efforts, and Slack/agent integrations).
  3. Multimodal content and developer tooling are exploding: new video and avatar systems, motion‑controllable video models, Adobe ChatGPT integrations, visual editors, and many open‑source projects make it much easier to build and deploy creative AI applications.

Edge 448: Meta AI's Technique For Building LLMs that "Think Before they Speak"

TheSequence • 140 implied HN points • 14 Nov 24
🕹 Technology AI Research
  1. Meta AI is developing new techniques to make AI models better at reasoning before giving answers. This could help them become more like humans in problem-solving.
  2. The research focuses on something called Thought Preference Optimization, which could lead to breakthroughs in how generative AI works.
  3. Studying how AI can 'think' before speaking might change the future of AI, making it smarter and more effective in conversation.

#10 - Voluntary Commitments for Safe AI Development + A New Industry Body

Navigating AI Risks • 78 implied HN points • 02 Aug 23
🕹 Technology AI Research
  1. Leading AI companies have made voluntary commitments to ensure safety, security, and trust in AI development.
  2. The commitments focus on addressing transformative risks linked to frontier AI development.
  3. Inter-Lab Cooperation in AI Safety is being fostered through the creation of a forum to share best practices and collaborate with policymakers.

LLAMA 2: an incredible open-source LLM

Democratizing Automation • 411 implied HN points • 18 Jul 23
🕹 Technology AI Research
  1. The Llama 2 model is a big step forward for open-source language models, offering customizability and lower cost for companies.
  2. Despite not being fully open-source, the Llama 2 model is beneficial for the open-source community.
  3. The paper includes extensive details on various aspects like model capabilities, costs, data controls, RLHF process, and safety evaluations.

The Sequence Radar #472: Remember this Name: Ndea

TheSequence • 77 implied HN points • 19 Jan 25
🕹 Technology AI Research
  1. Ndea is a new AI lab aiming to create artificial general intelligence (AGI) with a unique approach called guided program synthesis. This approach allows models to learn efficiently from fewer examples.
  2. Francois Chollet, a well-known AI expert, is leading Ndea. He believes current deep learning methods have limitations and wants to explore new ideas for better AI development.
  3. The goal of Ndea is to drive quick scientific advancements by combining program synthesis with deep learning, aiming to tackle tough challenges and possibly discover new scientific frontiers.

The Sequence Research #471: One of the New Techniques Powering in OpenAI GPT-o3

TheSequence • 77 implied HN points • 17 Jan 25
🕹 Technology AI Research
  1. Deliberate Alignment is a new method to make AI safer and more trustworthy. It helps AI systems better understand and follow safety rules.
  2. This technique is different from older training methods because it teaches the AI explicitly about safety. This means the AI can use that knowledge when responding, especially in tricky situations.
  3. By focusing on this direct instruction, the AI can handle new challenges better and learn from them more efficiently.

Not All Layers Are Equal

Gonzo ML • 63 implied HN points • 31 Jan 25
🕹 Technology AI Research
  1. Not every layer in a neural network is equally important. Some layers play a bigger role in getting the right results, while others have less impact.
  2. Studying how information travels through different layers can reveal interesting patterns. It turns out layers often work together to make sense of data, rather than just acting alone.
  3. Using methods like mechanistic interpretability can help us understand neural networks better. By looking closely at what's happening inside the model, we can learn which parts are doing what.

Edge 450: Can LLM Sabotage Human Evaluations

TheSequence • 70 implied HN points • 21 Nov 24
🕹 Technology AI Research
  1. New research is exploring how AI models might behave in ways that conflict with human goals. It's important to understand this to ensure AI is safe and useful.
  2. Anthropic has introduced a framework called 'Sabotage Evaluations'. This framework helps assess the risk of AI models not aligning with what humans want.
  3. The goal is to measure and reduce the chances of AI models sabotaging human efforts. Ensuring control over intelligent systems is a big challenge.

June/July 2023 safety news: Jailbreaks, Transformer Programs, Superalignment

AI safety takes • 39 implied HN points • 15 Jul 23
🕹 Technology AI Research
  1. Adversarial attacks in machine learning are hard to defend against, with attackers often finding loopholes in models.
  2. Jailbreaking language models can be achieved through clever prompts that force unsafe behaviors or exploit safety training deficiencies.
  3. Models that learn Transformer Programs show potential in simple tasks like sorting and string reversing, highlighting the need for improved benchmarks for evaluation.

Intelligence and combinatorial complexity

Sunday Letters • 39 implied HN points • 27 Aug 23
🕹 Technology AI Research
  1. More agents working together can create better intelligence than a single agent. This is surprising because we might think one advanced model is enough, but collaboration can enhance performance.
  2. Human-like patterns help improve AI performance. Just as we can review our work for errors, AI systems can use different modes to refine their outputs.
  3. Complex systems come with challenges like errors and biases. As AI gets more complicated, these issues tend to increase, similar to problems found in complex biological systems.