Figma bridged front-end engineering logic and design thinking for one practitioner, making structural relationships — components, styles, and tokens — tangible and learnable through hands-on friction. AI tooling now risks smoothing away exactly that friction, producing designers who generate plausible-looking work without deeply understanding the decisions behind it. Whether today's practitioners will develop equivalent structural intuition through different feedback loops remains an open question the industry has yet to answer.
2026.06.18·blog.murphytrueman.com·4 min read·TLDR Design LinkedIn's design community is increasingly criticized for favoring exaggerated claims, recycled AI commentary, and dramatic storytelling over clear, concise communication. Many common phrases and formatting habits create a false sense of urgency or insight, turning straightforward observations into attention-seeking content that often adds more noise than value.
Create premium, highly customizable videos, visuals, and 3D mockups that showcase UI designs, websites, and apps.
Moodboard 3000 is an image layout engine for Figma. Select images on canvas, choose a layout, and generate a composed moodboard.
Lunagraph lets you design and create UI using real HTML, CSS, and React code. Stay consistent with zero handoff and work between teams, designers, product, developers, and agents.
Designing with AI requires embracing probabilistic thinking — treating model outputs as weighted signals rather than definitive answers, since AI systems predict based on historical patterns that may embed biases and miss future shifts. Practical applications range from calibrating design effort to confidence levels to keeping humans in the loop for review, correction, and model improvement. Ultimately, the mindset shift is less about tools and more about framing: instead of asking, "Will this work?", designers should ask, "How likely is this to work, and what happens when it doesn't?"
2026.06.18·smashingmagazine.com·19 min read·TLDR Design Most AI marketing challenges — search visibility, recommendations, consistency, attention — are reframings of long-standing problems marketers have navigated before. As consumers delegate purchasing decisions to AI agents, algorithms rather than humans will become the target audience. Until that AI-to-AI marketplace materializes, today's challenges remain identical in nature to those marketers have already solved.
2026.06.18·brandingstrategyinsider.com·5 min read·TLDR Design Yuya Zhou's Genesis, a series of seven bioplastic light objects shown at Milan Design Week 2026, challenges the conventional permanence of design by using materials — starch, gelatin, glycerin, and edible pigments — that evolve, deform, and change over time.
Optical illusions reveal that the brain doesn't passively receive sensory data — it actively constructs perception, sometimes filling in gaps or generating images that aren't there.
Hannah Li's personal series, The Way Back, uses light, atmosphere, and everyday spaces to capture quiet moments of reflection, memory, and transition.
Sponsored
Same input. Same prompt. Different output. That's the reality of testing AI agents that write code, and most teams are shipping without solving it.
2026.06.18·links.tldrnewsletter.com·Sponsor·TLDR Dev 2026.06.18·links.tldrnewsletter.com·TLDR Dev A clear division of labor exists in agentic coding where humans make the majority of planning decisions while the AI handles technical execution. Success in these tasks depends more on the user's domain expertise than their formal programming skills, allowing professionals in various non-technical fields to get results comparable to those of software engineers. Recent trends show a shift toward complex, end-to-end agentic work and a measurable increase in the estimated economic value of the tasks being completed.
2026.06.18·anthropic.com·25 min read·TLDR Dev Dirge is a highly efficient coding harness developed in Rust that improves budget model performance while still having a minimal memory footprint. Unlike traditional harnesses that focus solely on model capabilities, Dirge's supportive framework, with features like error correction, session continuity, and a memory management system help optimize model quality across different contexts.
2026.06.18·yogthos.net·18 min read·TLDR Dev Browser Use's cloud browsers have been optimized to start quickly, remain isolated, and maintain low costs, now operating at $0.02 per hour with sessions starting in under a second. This was done by using Firecracker for lightweight virtual machines while implementing a control plane for dynamic scaling.
2026.06.18·browser-use.com·17 min read·TLDR Dev Conversations with others often lead to greater problem-solving and understanding than solitary thinking can provide as they introduce a feedback loop that refines thoughts in real-time.
2026.06.18·thesignalist.io·6 min read·TLDR Dev Yak shaving refers to the process of engaging in a series of tangential tasks that divert from the original goal, showing how building projects from scratch can lead to unexpected challenges, discoveries, and joy.
2026.06.18·parksb.github.io·8 min read·TLDR Dev Sponsored
The infra. Token metering, model routing, usage variance. Vercel, Replit, Glean, and Supabase use Orb [14] to handle it. See how 80 AI companies actually price in the 2026 State of AI agent pricing report [13].
eve is an open-source agent framework that makes building, running, and scaling agents easier by providing built-in production capabilities like durable execution, sandboxed computing, and human-in-the-loop approvals. With a clear structure, agents can be defined through a directory of files that dictate their components, actions, and connections.
2026.06.18·vercel.com·12 min read·TLDR Dev Lore is an open source version control system designed for scalability, particularly for projects with large binary assets like games. It has a centralized, content-addressed architecture that allows for efficient data storage and retrieval, with an emphasis on immutable revisions and low-overhead branching.
2026.06.18·github.com·GitHub repository·TLDR Dev TREX is a code review tool that integrates code execution with AI-driven testing to identify bugs that may not be apparent in static analysis alone. It uses a multi-agent architecture, where a Greptile reviewer orchestrates specialized TREX agents that run tests in isolated environments, providing artifacts such as screenshots and logs to verify findings.
2026.06.18·greptile.com·9 min read·TLDR Dev Image compression techniques, including lossy and lossless methods, exploit characteristics of human visual perception to reduce file sizes. Formats like JPEG use a combination of chroma subsampling, discrete cosine transforms, and quantization for compression without noticeable quality loss, while GIF uses indexed color palettes and Lempel-Ziv-Welch algorithms to store limited color data.
2026.06.18·makingsoftware.com·35 min read·TLDR Dev An analysis of various LLMs in triaging security vulnerabilities showed that a council of four LLMs achieved an impressive 86.2% unanimous agreement on findings, while lower reasoning efforts consistently performed poorly. Although higher reasoning generally correlated with better results, there were exceptions where models struggled with complex tasks.
2026.06.18·parsiya.net·26 min read·TLDR Dev Sponsored
Foxit's open-source MCP server gives AI agents access to PDF conversion, OCR, merge, and signing. Works with Claude Desktop, Cursor, and GitHub Copilot. Set it up in under 15 minutes [20].
2026.06.18·developer-api.foxit.com·Sponsor·TLDR Dev Anthropic has released a playbook of practical exercises and agentic workflows to navigate every stage of the business lifecycle for AI-native startups.
2026.06.18·claude.com·5 min read·TLDR Dev mac-ocr is a macOS command-line tool that uses Apple's Vision framework locally to recognize text from images and PDFs and create searchable PDFs.
2026.06.18·github.com·GitHub repository·TLDR Dev Gzip can be used as a language model through compression techniques that predict and generate text continuations by scoring potential outputs based on how well they compress with a given context.
2026.06.18·nathan.rs·5 min read·TLDR Dev Exa Agent is a new API that combines frontier language models with optimized web search tools to provide high-performance research, list-building, and data enrichment capabilities.
LTK led share of voice in YouTube's GRWM videos, despite only appearing in descriptions via affiliate links. The format is largely unpaid, with nearly 90% of mentions driven by creator choice. Prestige brands lost ground while Sephora gained share, and budget brands surged. Rising platforms like ShopMy also point to increasing competition in the creator monetization layer.
ChatGPT's product recommendations depend heavily on whether search is enabled, which means brands should focus less on influencing AI training data and more on influencing the sources AI cites. In a study of 20,000 responses, 80.2% of recommended products changed when search was enabled, and only 15.8% of products that appeared in 100% of non-search responses also appeared when search was enabled. The findings suggest that cited web content plays a major role in recommendation outcomes. Brands should focus on earning mentions in the articles, reviews, and comparison pages that ChatGPT already cites for high-value product queries.
Paid media now plays a critical role in AI visibility because it helps you identify winning messages and audience signals faster than any other channel. High-performing ad copy should inform content, community engagement, and thought leadership since the language that drives clicks often becomes the language buyers recognize across Reddit, YouTube, review sites, and other sources AI models cite. Paid media should also fund distribution, since content that reaches and influences audiences is more likely to generate the third-party mentions and citations that shape AI answers. Treat paid, organic, SEO, and community channels as a connected system where paid uncovers insights, validates positioning, and accelerates the spread of content that can compound visibility over time.
Most B2B LinkedIn content fails at one of two extremes: too dense for a social feed or too shallow to earn professional credibility. The 95/5 playbook addresses this by structuring content across three complexity levels mapped to where a buyer sits in their decision journey, then pairing it with two synchronized motions. Zero-click thought leadership builds familiarity with the passive 95%, and signal-triggered outbound converts the active 5%. High-intent signals like gated downloads or pricing page visits earn a direct message referencing the exact resource. Low-intent signals like likes and comments get a softer reply first.
Widgets let users create and edit interactive email elements through chat instead of code. After adding a widget like a quiz or scratch card, users can describe changes in plain language. The assistant handles configuration, generates HTML, AMP, and fallbacks. It also applies updates instantly, reducing the effort required to build and maintain interactive emails.
Soundcore Work is a clip-on AI recorder designed for in-person meetings. It records conversations, flags key moments with a tap, and syncs transcripts to a phone in real time. Afterward, it generates a one-page summary and lets users query the transcript for specific details, reducing the need for manual note-taking.
Hinge is building brand credibility with Gen Z by funding the social infrastructure its users depend on rather than treating community as a marketing channel. The ‘One More Hour' initiative has distributed more than $2M in grants to grassroots social groups across New York, Los Angeles, Atlanta, and London, with $1.5M more planned globally this year, requiring no promotion or app downloads from recipients. The strategy is a deliberate contrast to brands pulling back from purpose-driven marketing, grounded in a real behavioral gap where Gen Z spends 1,000 fewer hours of in-person connection per year than peers did 20 years ago.
Granola grew by removing friction instead of adding features. It avoided meeting bots and recorded audio at the operating system level, which let it work across Zoom, Meet, Teams, phone calls, and other audio sources without integrations. Its note taking workflow increased trust by letting users create lightweight notes that AI expands into useful summaries, which made the output feel personal and worth revisiting. Granola also built habit loops through simple product design such as one click meeting notifications that launch the call, open the app, and start transcription. After earning trust as a notetaker, it turned meeting data into a searchable knowledge base through chat, shared team spaces, and reusable prompts, which expanded its value beyond note taking and helped drive word of mouth growth, 70%+ weekly retention, and a $1.5B valuation.
Consumers associate fewer ingredients with products that feel more natural, healthier, and less processed, making ingredient simplicity a strong messaging lever for food and beverage brands. Across 19 experiments, framing a product as having "few ingredients" increased choice rates by 21.6% for an identical granola bar and generated up to 44.2% higher click-through rates in Meta ads. Brands should prominently highlight low ingredient counts on packaging, ads, and product pages, especially for health-focused products where differentiation is difficult. The tactic is less effective for products bought for indulgence, flavor variety, or nutritional complexity, so brands should match the message to the purchase motivation.
TikTok has added new rules that ban the use of AI-generated voices or recordings in live broadcasts.
Social media has overtaken TV and news websites as the most-used daily news source, with TikTok and Instagram rising in influence as X continues to lose ground.
Oversaturating your feeds with content may be a smart way to break through vs. scaling back.
Sponsored
Your team is using AI as a productivity hack. But for real ROI, there's something missing: business context.
2026.06.17·campaigns.celonis.com·Sponsor·TLDR 2026.06.17·campaigns.celonis.com·TLDR SpaceX is buying AI startup Cursor for $60 billion worth of stock. Cursor, which has a popular AI coding tool that has experienced explosive growth since the company's founding in 2022, crossed $1 billion in annualized revenue in November. SpaceX has yet to provide investors with details on Cursor's customer list, momentum, or revenue. The merger is expected to close during the third quarter, subject to requisite regulatory approvals.
Apple's camera-equipped AirPods are scheduled to launch in late 2027. The device will be released around the same time as the foldable iPhone and a 20th anniversary iPhone model. All three products have reached advanced stages of development. Apple intends the release to be its biggest wave of new products yet.
2026.06.17·links.tldrnewsletter.com·5 min read·TLDR Genesis AI, a startup backed by former Google CEO Eric Schmidt, has unveiled a general-purpose robot called Eno. The robot can reason, adapt, and own outcomes beyond predefined tasks. Genesis AI is working with LG Group's consulting and services arm to deploy the robot to industrial customers by the end of the year. The startup is currently raising funds to finance its next steps.
2026.06.17·links.tldrnewsletter.com·3 min read·TLDR Snap's Specs, at $2,195 with a $200 refundable deposit, are expected to ship later this year in the US, UK, and France. The augmented reality glasses bring digital visuals into a user's field of vision, creating a new way to use computing in shared experiences in the real world. Snap's earlier attempt at a similar product, the $130 camera only Spectacles that debuted in 2016, never became a hit. Meta has more recently found some success with its Ray-Ban Meta glasses, and Google plans to launch AI-powered glasses developed together with Samsung.
Sponsored
Stop stitching together CDC pipelines and managing ETL jobs. Just build on Lakebase.
2026.06.17·links.tldrnewsletter.com·Sponsor·TLDR AWS has a new metadata capability for Amazon Simple Storage Service (S3) that enables users to attach rich, large-scale business context directly to their objects. The annotations feature allows users to store up to 1,000 named annotations per object, each up to 1 MB in size, in flexible formats like JSON, XML, YAML, or plain text. Annotations can be modified or deleted at any time without rewriting objects. The feature is available now in all AWS regions.
2026.06.17·aws.amazon.com·9 min read·TLDR Meta has been known over the last two decades for being a unique, high-performance engineering organization. It had a 'move-fast-and-break-things' culture where engineers were empowered to do good work, focus on impact, and balance business interests with solid engineering. This has all changed in the past few weeks. This post looks at how bad things have become since April, what's happened, what's going through the minds of leadership, and the other factors causing the organization to demolish a proven, successful engineering culture in the most ruthlessly efficient way possible.
2026.06.17·newsletter.pragmaticengineer.com·30 min read·TLDR Audited financial statements obtained by an independent journalist show that OpenAI's reported revenue grew from $3.7 billion in 2024 to $13.07 billion in 2025. However, its expenses grew from $7.81 billion in 2024 to $19.18 billion in 2025. These numbers reflect the high costs OpenAI has incurred in training new models. The losses point to a concerning direction for a company that is telling investors that it hopes to be profitable by 2030.
2026.06.17·arstechnica.com·4 min read·TLDR