Search the archive

THE JOYS (AND ABSURDITIES) OF AI SLOP

Brands can grab attention with absurd AI content built around a clear idea.

THE MONEY INFLUENCERS ACTUALLY MAKE

A look into how much 26 Boston-area content creators earn from paid posts, UGC, and affiliate links.

7 TIPS TO WIN ON REDDIT

Reddit tactics that focus on remarketing ads, active subreddits, user reviews, and AMAs.

See what's possible

ALIBABA'S MAIN AI APP DEBUTS STRONGLY IN EFFORT TO RIVAL CHATGPT

Alibaba's Qwen app drew more than 10 million downloads in the week after its relaunch. The company will gradually add agentic AI features to support online shopping in the coming months. Alibaba has rebranded itself as an AI-first business. It plans to deeply integrate core lifestyle and productivity services directly into the Qwen app.

ELON MUSK'S BORING COMPANY REVEALS PRUFROCK TBM'S MOST DISRUPTIVE FEATURE

The Boring Company's Prufrock-5 tunnel boring machine (TBM) will begin tunneling by December 1. One of the key features of the Prufrock machines is that they can be retrieved, upgraded, and deployed again with improvements. Standard TBMs are often left underground at the end of projects because retrieving them is usually too expensive and impractical. TBMs from the Boring Company could eventually be as reusable as SpaceX's boosters.

MAPPING THE SENSE OF WHAT'S GOING ON INSIDE

Interoception is a sense that supplies the brain with a remarkably rich picture of what is happening throughout the body. It shapes our emotions, behavior, decisions, and feelings. A growing body of research suggests that many psychiatric conditions may be caused in part by errors in our perception of our internal environment. Scientists still need to develop a firm understanding of how interoception works, but someday it may be possible to treat some conditions by returning a person's internal sense.

BRAIN HAS FIVE ‘ERAS', SCIENTISTS SAY – WITH ADULT MODE NOT STARTING UNTIL EARLY 30S

The brain experiences five major phases, split up by four pivotal turning points (at around the ages of nine, 32, 66, and 83 years) in which brain organization moves on to a different trajectory. Understanding the brain's structural journey helps scientists identify when and how its wiring is vulnerable to disruption. It could also give insights into risk factors for mental health disorders, which most frequently emerge during the adolescent period. An overview of the five stages is available in the article.

learning. Download the guide

NEW GEMINI API UPDATES FOR GEMINI 3

Gemini 3 is now available for developers to build with via the Gemini API. Google has rolled out several updates to the API to support Gemini 3's state-of-the-art reasoning, autonomous coding, multimodal reasoning, and agentic capabilities. The changes are designed to give users more control over how the model reasons, how it processes media, and how it interacts with the outside world. Details are available in the article.

WHAT A CTO SHOULD KNOW ABOUT TECH

Tech leaders should take the time to learn what they need to know about tech to be effective. Their job as a CTO is to make decisions in an extremely complicated environment and under conditions of considerable uncertainty. It's really important to understand clearly what tools and what teams are working with. A lack of understanding will eventually cause a messy crash.

TECH PREDICTIONS FOR 2026 AND BEYOND

Technology has become so intertwined with day-to-day life that it influences everything, from our relationships to the way we work, protect ourselves, and even the things we choose to learn and when. The next year will see humans transition into a new era of AI in the human loop. This cycle will create massive opportunities to solve problems that truly matter.

THE MATH OF WHY YOU CAN'T FOCUS AT WORK

You are not uniquely undisciplined. Deep work is mathematically rare unless you deliberately design for it. Just one less interruption per hour each day can transform your week. Learn how to split up tasks into smaller chunks. This post presents a model for increasing focus at work.

Subscribe now.

WE NEARLY HAD POWER PROFILING IN CHROMIUM

Power profiling in Firefox allows developers to get granular information about the energy being consumed by the browser as a website is used.

NVIDIA CEO TELLS EMPLOYEES TO USE MORE AI — AND DON'T WORRY ABOUT THEIR JOBS

Nvidia CEO Jensen Huang wants every task that can be automated with AI to be automated with AI.

WEBGPU IS NOW SUPPORTED IN MAJOR BROWSERS

WebGPU, a powerful API for running high-performance 3D graphics and general-purpose GPU computations, is now officially supported in Chrome, Edge, Firefox, and Safari.

NVIDIA IS SO SPOOKED BY GOOGLE'S SUDDEN AI COMEBACK THAT IT'S POSTING ON X TO DEFEND ITSELF

Nvidia posted on X to defend itself after a report suggested Meta was considering shifting part of its AI infrastructure to Google's TPUs.

WHAT THEY DON'T TELL YOU ABOUT MAINTAINING AN OPEN SOURCE PROJECT

Shipping is just the beginning.

MEMORIES OF .US

A look at the history of the .us domain and why it failed to take off.

Reserve your spot

MUSK'S XAI TO CLOSE $15 BILLION FUNDING ROUND IN DECEMBER

Elon Musk's xAI plans to close a $15 billion funding round with a $230 billion pre-money valuation in December. This comes months after similar large raises by OpenAI and Anthropic.

AMAZON TO SPEND UP TO $50 BILLION ON AI INFRASTRUCTURE FOR THE US GOVERNMENT

Amazon will invest up to $50 billion to expand its capacity to provide AI and high-performance capabilities for its cloud unit's US government customers. The project will begin in 2026 and add nearly 1.3 gigawatts of capacity. It will give government agents access to AWS' AI tools, Anthropic's Claude family of models, Nvidia chips, and Amazon's custom Trainium AI chips. The project will enable agencies to develop custom AI solutions, optimize data sets, and enhance workforce productivity.

ILYA SUTSKEVER – WE'RE MOVING FROM THE AGE OF SCALING TO THE AGE OF RESEARCH

This post contains a transcript of an interview with Ilya Sutskever, former chief scientist at OpenAI and co-founder of Safe Superintelligence. In the interview, he discusses model jaggedness, emotions and value functions, why humans generalize better than models, alignment, and more. Sutskever says that current models generalize dramatically worse than people and that it is a very fundamental thing. Scaling has gotten big - it's time to get back to research to continue creating improvements.

LLM | UNIT ECONOMICS

OpenAI and Anthropic likely won't ever stop training entirely. However, they don't need to grow training spend by multiples forever. The moment annual training spend stops growing 5x a year, profit margins will show up almost immediately. These companies are currently burn machines, but they won't always be.

THE ECONOMICS OF REPLACING CALL CENTER WORKERS WITH AIS

AI voice agent rates are currently competitive with human labor in some countries, but it is still cheaper to hire humans in most developing countries. Inference costs are massively decreasing every year. Voice agents will likely be competitive with the world's cheapest human labor in around 2030.

NANO BANANA PRO: RAW INTELLIGENCE WITH TOOL USE

Nano Banana Pro has pushed the frontier of infographic generation. It pulls data and synthesizes it to get the best results. The model can be prompt engineered for extremely nuanced AI image generation. Examples of outputs generated by the model are available in the article.

cloud controller. Try it for free

CHATGPT 5.1 CODEX MAX

OpenAI's GPT-5.1-Codex-Max enhances capabilities over GPT-5.1-Codex, showing improved performance in SWE-bench-verified, SWE-Lancer-IC SWE, and Terminal-Bench 2.0. It advances task persistence, cybersecurity preparations, and introduces Windows training, while network access remains disabled by default for security. Despite substantial internal evaluation progress, external reviews show mixed cybersecurity capabilities, yet highlight advances in AI self-improvement tasks and overall code efficiency.

FLUX.2 RELEASE

The newly released FLUX.2 is designed for real-world creative workflows. It produces consistent, high-quality images from structured prompts and multiple references. The model handles brand constraints, lighting, and layout with editing support up to 4 megapixels.

GOOGLE ANTIGRAVITY EXFILTRATES DATA

Google's Antigravity coding tool can be manipulated through hidden instructions in poisoned web pages (aka prompt injection). A fake Oracle integration guide was used to trick Gemini into bypassing its own security settings, collect user credentials from protected files, and send the stolen data to an attacker-controlled website included in Antigravity's default allowed domains list.

SUNO CREATES AN ENTIRE SPOTIFY CATALOG'S WORTH OF MUSIC EVERY TWO WEEKS, SAYS INVESTOR PITCH DECK FOR $250M FUNDRAISE

Users on the AI music platform Suno create as much music every two weeks as what is currently available on Spotify. Suno aims to expand its offerings so that eventually, creators and listeners will not need to leave its app to create, stream, or share their music socially. The company recently raised $250 million in a series C fundraise that brought its valuation up to $2.45 billion. The funding will help the company empower more artists to experiment, collaborate, and build on their creativity.

CHARACTER.AI INTRODUCES STORIES FOR INTERACTIVE FICTION

Character.AI's Stories is a visual and replayable interactive fiction format where users select Characters, choose a genre, and guide branching narratives by making decisions.

ELON MUSK PROPOSES GROK 5 VS WORLD'S BEST LEAGUE OF LEGENDS TEAM MATCH

No match has been confirmed, but the interest from players, teams, and Riot suggests that the concept could materialize into a landmark AI-versus-human matchup.

NVIDIA SAYS ITS GPUS ARE A 'GENERATION AHEAD' OF GOOGLE'S AI CHIPS

Nvidia shares fell 3% after reports that Meta may strike a deal with Google to use its AI chips.

NOW YOU CAN USE CHATGPT VOICE WITHOUT LEAVING YOUR CHAT

Users can now interact with ChatGPT Voice directly within ongoing text chats by clicking on the waveform icon next to the text field.

PERPLEXITY'S GLOBAL APP DOWNLOADS HAVE DROPPED 80% IN THE LAST SIX WEEKS

This implies that the app's earlier growth was mostly from paid marketing.

LLMS CAN INVENT THEIR OWN COMPRESSION

Models can discover their own compression tricks to pack more information per token.

BITCOIN REBOUNDS ABOVE $88K

Bitcoin recovered to $87,600, up 2.5% from its $85,550 weekend low. There was a shift from aggressive selling to orderly derisking with stabilizing open interest and muted spot activity despite persistent ETF outflows. Options sentiment turned cautiously optimistic as the 25-delta skew rebounded from -10.96 to -4.58 while Long Call Condor block trades across $100,000-$118,000 strikes indicate demand for upside exposure into late 2025. Analysts believe Bitcoin must reclaim the $87,000-$88,000 range for sustained gains. Institutions are patiently bullish and building conviction ahead of December's Federal Reserve interest rate decision, which may be the next major catalyst.

CHINA RETURNS AS BITCOIN MINING HUB WITH 14% MARKET SHARE

China reclaimed the third-largest Bitcoin mining position with approximately 14% of global hashrate by October despite the 2021 ban, driven by cheap electricity and excess data center capacity in Xinjiang and Sichuan provinces. 15-20% of global mining capacity now operates underground in China, which has softer policy signals, including Hong Kong's stablecoin legislation and yuan-backed stablecoin discussions indicating a more flexible stance. Bitcoin hashprice hit an all-time low of $34.2 per petahash per second on Friday, driven by Bitcoin's 30%+ decline from October peaks, subdued transaction fees, and network hashrate above one zettahash (10% below recent highs). The next difficulty adjustment is expected on Wednesday.

FULL SEND: FASTEST PATH FOR TRANSACTION INCLUSION

CROSS RIVER LAUNCHES STABLECOIN PAYMENTS

Cross River Bank launched a stablecoin payments infrastructure, unifying fiat and stablecoin flows in a single interoperable system that eliminates pre-funding requirements and ledger rebuilding for network settlement, merchant payouts, and treasury management. The platform targets $20 trillion in annual stablecoin volumes by providing bank-grade compliance and programmability for fintechs, enterprises, and crypto-native companies without operational sprawl across fragmented rails. The infrastructure is available today for approved partners, with broader availability planned.

UNDERSTANDING SIMD-0411

Solana's SIMD-0411 proposal doubles the disinflation rate from -15% to -30% annually, accelerating terminal inflation arrival from 6.2 years to 3.1 years while reducing emissions by 22.3 million SOL over six years, worth $2.9 billion at current prices. The single-parameter change delivers similar emissions reduction as the failed SIMD-0228 without introducing dynamic, stake-responsive formulas that raised implementation complexity and governance friction. Nominal staking yields decline from the current 4.18% to 5.04% in year one, 3.48% in year two, and 2.42% in year three, pushing approximately 10 validators into unprofitability initially while strengthening long-term SOL value through reduced dilution. The deterministic, easily modeled approach contrasts with SIMD-0228's exponential function, tying inflation to staking participation percentages, offering validators and institutions predictable outcomes without changing the 1.5% terminal rate the ecosystem already aligns around.

COINBASE VENTURES: WHAT THEY WANT TO FUND IN 2026

TOKENIZATION ONLY MATTERS IF MONEY MOVES

Tokenizing an asset is just the starting point, as without either long-term lock-in or frequent real usage, you've just created “dead money” that sits onchain and earns nothing. There are only two long-lasting ways for a tokenized asset to generate fees: it must attract committed, parked capital by paying yield and rewarding conviction, or be used constantly so velocity plus fees make it valuable.