In this webinar, the host will show how to use his "research agent stack". It deploys Deep Research to gather massive amounts of web data, then funnels it into NotebookLM to create a hallucination-free knowledge base. The process allows you to turn thousands of pages of research into instant briefings, podcasts, or "voice style guides" in one click. The free event will take place on February 13 at 9 AM PT.
This guide teaches how to use Claude Opus 4.6 to automate workflows by combining Skills, Cowork, and Plugins. Skills let you document processes once and have Claude follow them consistently. Cowork connects Claude to files, apps, and multi-step tasks, while Plugins bundle Skills with tool integrations and slash commands. The guide covers building Skills without code, applying them across content and visualization tasks, leveraging pre-built libraries, and creating custom plugins to scale team output efficiently.
Experiments let advertisers test PMax against other campaign types, final URL expansion, or measure incremental gains with Uplift tests. Data Exclusions now allows blocking specific audiences, including customer match and remarketing lists, from seeing ads. Product Overlap identifies Shopping campaign conflicts, helping advertisers optimize performance and exclude underperforming products.
Growth engineers are a hybrid of growth marketer and engineer. They think less about running experiments and more about building the infrastructure that enables those experiments. AI tools enable teams to execute independently, accelerating iteration as software build costs fall. Canva's programmatic SEO templates illustrate this scale, and companies like Ramp are already hiring for this hybrid role.
In 2011, Facebook's Open Graph let apps automatically share user activity on friends' feeds. Spotify leveraged this by connecting listens to Facebook, turning each play into a discovery loop that drove millions of new users. It increased engagement and boosted paid subscriptions fivefold in under two years. Many apps that relied solely on virality, like SocialCam and Viddy, collapsed when Facebook limited auto-sharing. Sustainable growth depends on both platform tailwinds and core product value.
A new Claude connector lets site owners securely share backend CMS data with read only access so Claude can analyze content without editing anything.
47% of marketers face AI errors weekly, and over 70% spend 1-5 hours fact-checking.
Anthropic's latest version of Opus features Agent Teams, teams of agents that can split larger tasks into segmented jobs. It also comes with a one-million-token context window, allowing for the processing of larger documents. The segmenting of agent responsibilities allows them to coordinate work in parallel to complete tasks faster. Agent Teams is currently available in a research preview for API users and subscribers.
2026.02.06·techcrunch.com·2 min read·TLDR OpenAI Frontier is a new AI platform for building, deploying, and overseeing AI agents. It makes it easier for businesses to combine sources of data that agents need to perform tasks. The agents act as co-workers that can collaborate with humans and can be used alongside agents developed by other labs. Frontier is currently available to a limited set of customers. OpenAI has yet to disclose prices for the feature.
2026.02.06·links.tldrnewsletter.com·3 min read·TLDR SpaceX and xAI signed a merger agreement at the end of last month. Merging the two companies is a level of risk-taking and ambition that is novel even for Elon Musk. It is a bet on Musk's plan for the convergence of AI and space - a wager on unproven technology deployed at a vast scale. SpaceX representatives have confirmed that the company plans to go public this summer.
2026.02.06·links.tldrnewsletter.com·8 min read·TLDR Nearly 90% of all humanoid robots sold globally last year were Chinese. The humanoid robot industry is still in its infancy, but China's progress will potentially lock it into a commanding position. Its early advantage is due to a combination of policy support, public investment, a mature supply chain, and advancements made in AI software and hardware. The global market for humanoid robots in 2035 is estimated to grow to $38 billion.
2026.02.06·restofworld.org·5 min read·TLDR GPT-5.3-Codex is OpenAI's most capable agentic coding model to date. It advances both the frontier coding performance of GPT-5.2-Codex and the reasoning and professional knowledge capabilities of GPT-5.2 while being 25% faster. The model can take on long-running tasks that involve research, tool use, and complex execution. Developers can steer and interact with the model while it's working without losing context.
2026.02.06·links.tldrnewsletter.com·12 min read·TLDR Mitchell Hashimoto, a developer who co-founded HashiCorp, has reached a point where he is having success with modern AI tooling. His approach uses a proper, measured view grounded in reality. This post shares Hashimoto's approach to navigating these new tools and gives a glimpse into how he approaches new tools in general, regardless of AI. Hashimoto is a software craftsman who just wants to build stuff for the love of the game, so it doesn't matter to him whether AI is here to stay.
2026.02.06·mitchellh.com·14 min read·TLDR Google is expanding AirDrop support to the rest of the Android ecosystem this year. Currently, only Google Pixel 10 devices can initiate an AirDrop session with Apple devices. AirDrop support allows Android devices to send files to Apple devices without downloading third-party apps. Google tends to time its Android ecosystem updates with Pixel Drops, and the next one of those is expected in March.
2026.02.06·arstechnica.com·2 min read·TLDR Wall Street tends to run in herds, and the past couple of weeks have seen the herd collectively conclude that software is somehow dead. This is nonsense. There will be more software than ever before. AI-enabled or AI-centric software is simply moving up the stack of what a product is. Domain experience will become more important than it is today because every domain will become vastly more sophisticated than it is now.
2026.02.06·links.tldrnewsletter.com·13 min read·TLDR Apple glasses are rumored to be unveiled this year, and the new Apple Health+ service is expected to arrive in iOS 27 with AI at the center.
2026.02.06·9to5mac.com·2 min read·TLDR AI enables developers to start doing the projects they've been putting off for the future.
2026.02.06·ghuntley.com·5 min read·TLDR NASA Administrator Jared Isaacman has told his team to challenge NASA's bloated requirements to ensure they are still needed today.
2026.02.06·arstechnica.com·2 min read·TLDR Investors are worried that the hundreds of billions of dollars spent on AI data centers will take years to pay off.
2026.02.06·links.tldrnewsletter.com·7 min read·TLDR SpaceX advisers have reached out to major index providers to discuss whether SpaceX could join key indexes sooner than normal.
2026.02.06·links.tldrnewsletter.com·4 min read·TLDR Elon Musk plans to leverage xAI's ability to create and scale hardware to enable space-based data centers that will run on basically unlimited power from solar energy.
2026.02.06·dwarkesh.com·134 min read·TLDR Sponsored
Superagent is Airtable's new standalone product: an AI that deeply interrogates your business questions by building a wide-reaching research plan, deploying agents at scale to execute, and scouring credible sources to build a complete answer.
2026.02.06·superagent.com·Sponsor·TLDR AI OpenAI's GPT‑5.3‑Codex is a faster agentic coding model that combines GPT‑5.2‑Codex's coding performance with GPT‑5.2's reasoning and professional knowledge.
2026.02.06·links.tldrnewsletter.com·11 min read·TLDR AI Anthropic has released Claude Opus 4.6, an upgraded flagship model with stronger agentic coding, longer task persistence, and improved performance in large codebases. The model introduced a 1M-token context window in beta and achieved state-of-the-art results across several reasoning, coding, and economically valuable work evaluations.
2026.02.06·anthropic.com·12 min read·TLDR AI OpenAI desperately needs cash because it lacks economies of scale. Microsoft is shipping billions to OpenAI because it would be disastrous for the company if OpenAI went under. $10 billion of OpenAI's revenue may have been effectively a subsidy from Microsoft. This means that OpenAI was charging subscribers around 25% of what they were costing. OpenAI would have significantly fewer subscribers with an $80 per month subscription.
2026.02.06·blog.dshr.org·16 min read·TLDR AI Many people drastically underestimate what Codex can do. This post shares a simple setup that is routinely allocated hundreds of millions of tokens. The author spent $10,000 on API costs this month. The post also shares some killer use-cases for Codex, and reflects on how organizations might become significantly more efficient in the near future.
2026.02.06·links.tldrnewsletter.com·6 min read·TLDR AI Sponsored
Join CData and Microsoft on 2/18 for a live build of secure, cross-functional agentic workflows spanning CRM, ERP, and billing. Get Microsoft's best practices to build an agentic infrastructure and how CData Connect AI provides the connectivity, context, and control to
2026.02.06·cdata.com·Sponsor·TLDR AI The new v0 expands from demo tooling into a full production-ready platform with enterprise-grade security, integrations, and support for building and deploying real software with AI.
2026.02.06·vercel.com·4 min read·TLDR AI Multiple Claude instances worked in parallel to build a Rust-based C compiler capable of compiling the Linux 6.9 kernel. This approach reduced human supervision and explored autonomous LLM agent capabilities via stress tests and a $20,000, 2,000-session effort. The compiler, while functional, has limitations like inefficient code and reliance on GCC for some features.
2026.02.06·anthropic.com·13 min read·TLDR AI The automation of AI research and engineering will begin in earnest this year. This will create at least a step-change improvement in AI progress. It could also change the dynamics of AI competition, alter AI geopolitics, and much more. There will be more AI progress this year than in all of the years that have come before, and that's a conservative forecast.
2026.02.06·hyperdimensional.co·18 min read·TLDR AI Google's first-party models now process more than 10 billion tokens per minute via direct API usage. This is a 52x increase year-over-year. Google is committing capital at an unprecedented scale to fuel this growth. The level of investment signals the company's conviction that the demand for tokens is only just beginning.
2026.02.06·tomtunguz.com·3 min read·TLDR AI Sponsored
They're using SerpApi: the industry-standard Web Search API that shares access to search engines with a simple API. Trusted by Uber,
2026.02.06·serpapi.com·Sponsor·TLDR AI OpenAI Frontier is an enterprise platform for building, deploying, and managing AI agents across organizations.
2026.02.06·links.tldrnewsletter.com·8 min read·TLDR AI Turning technology proposals and architectural initiatives into shared transcripts captures the entire reasoning process, not just conclusions.
2026.02.06·davegriffith.substack.com·12 min read·TLDR AI The article discusses lessons from adopting AI tools in software development, emphasizing efficiency.
2026.02.06·mitchellh.com·12 min read·TLDR AI ElevenLabs CEO Mati Staniszewski claims voice is poised to become the primary interface for AI, surpassing text and screens.
2026.02.06·techcrunch.com·3 min read·TLDR AI Bitcoin tumbled more than 10% over 24 hours to $63,000 during early afternoon US trading Thursday, marking its steepest one-day drawdown since the FTX collapse on November 8, 2022 (14.3% drop) and dropping the largest cryptocurrency 50% below its record high above $126,000 hit in early October to its weakest level since October 2024. The selloff extended across asset classes with silver plunging 14% (now nearly 40% below its record from a week ago), gold falling over 2%, software stocks declining more than 3%, and crypto equities all tumbling over 10% alongside major crypto miners. Analysts attributed the severity to "very thin liquidity" where modest sell pressure triggers cascading liquidations in a fragile market environment with few orders to cushion trades, warning there is "still no signal that we bottomed out" and pointing to the 200-day moving average around $58,000 to $60,000 as the next key support level that aligns with bitcoin's average cost basis among all holders.
A coalition of European tokenization companies has urged EU policymakers to swiftly amend the bloc's DLT Pilot Regime, warning that current asset limits, volume caps, and six-year license restrictions are preventing regulated onchain markets from scaling while the United States advances toward industrial-scale tokenization and near-instant settlement. The companies proposed a narrow technical fix that would expand eligible assets, raise issuance caps, and remove pilot license time limits while maintaining investor protections, cautioning that the EU's broader Market Integration and Supervision Package won't take effect until at least 2030. The urgent call contrasts sharply with recent US regulatory momentum.
Payy Network is Ethereum's first privacy-enabled EVM L2, where ERC-20 transfers are private by default without requiring smart contract modifications or specialized wallets. The architecture uses a chain-native privacy pool for direct transfers and offchain "Privacy Vaults" that allow developers to configure privacy levels via RPC parameters, creating a spectrum from fully private to fully compliant/public. The network targets both crypto natives seeking frictionless private transfers and TradFi institutions wanting to bring flows onchain without front-running or MEV risks.
2026.02.06·links.tldrnewsletter.com·4 min read·TLDR Crypto Pandora, a decentralized prediction market platform that utilizes AI agents as resolution oracles, is deploying on Ethereum because of its proven anti-fragility, million-plus validator set providing the deepest security budget in smart contracts, and consistent upgrade execution as critical factors for infrastructure requiring credible neutrality and long-term reliability.
2026.02.06·links.tldrnewsletter.com·5 min read·TLDR Crypto Stablecoin profitability relies on issuance, flow, and payment acceptance. While issuance dominates today via interest on reserves, it remains cyclical and commoditized. Payment acceptance offers a sustainable alternative by leveraging interchange fees and organic liquidity, positioning future winners as payment platforms rather than simple issuers.
2026.02.06·links.tldrnewsletter.com·5 min read·TLDR Crypto Base is emerging as the dominant blockchain for AI agent activity due to transaction costs low enough to support micropayments via x402 and purpose-built infrastructure. The ecosystem has processed over 9M x402 transactions with $1M+ volume in 30 days, ~1.6M OpenClaw agents registered, and $100M+ volume on platforms like Bankr and Clanker. This represents one of the first demonstrable product-market fits for L2s beyond DeFi.
2026.02.06·links.tldrnewsletter.com·5 min read·TLDR Crypto Crypto entered a full winter in January 2025, with BTC down 39% and ETH down 53% from ATH, though institutional flows disguised the severity. ETF and DAT purchases totaling ~$75B propped up BTC/ETH/XRP (down 10-20%), while unsupported assets like ADA, AVAX, SUI, and DOT collapsed 62-75%. Retail crypto has been in a brutal winter that institutional capital merely papered over.
2026.02.06·links.tldrnewsletter.com·6 min read·TLDR Crypto Stablecoin payroll is scaling fast. Companies and payroll platforms are depositing real wages into wallets (embedded or self-custodial), creating recurring on-chain distribution that will only matter if people can actually spend those balances at checkout. WalletConnect Pay is presented as the missing acceptance layer. It's a PSP-grade integration that lets merchants and wallets use stablecoins like any other payment method so payroll-driven stablecoin volume can convert into everyday retail payments.
2026.02.06·links.tldrnewsletter.com·3 min read·TLDR Crypto Only 7 of 51 tokens finished green this week.
The value from stablecoins, L1s, and RWA tokenization isn't accruing to incumbent infrastructure because new entrants want to own the full stack, so investors should stop buying base layers and instead back application-layer winners.
A comparison of TVL versus borrowed capital shows real demand clustering around a handful of money markets.