AI Is Coming For Your Job

Good morning,

This week, I’m taking the lazy route. Or maybe the honest one. Either way, I’m not going to write much—because someone else already did.

Last week, Fiverr’s top dog, Micha Kaufman, sent a private memo to his employees. It was supposed to stay inside the building.

It didn’t.

And you need to read it.

Not because it’s dramatic. Not because it’s controversial. But because it’s true. Brutally, blazingly true. The kind of truth that leaves a ringing in your ears, like a gunshot just went off next to your cubicle.

Kaufman isn’t mincing words. He’s telling his people—designers, coders, support reps, lawyers, everyone—that AI is coming for their jobs. Not ten years from now. Not after another round of government panels and TED Talks. Now. This year. And if you’re not already sweating, you’re not paying attention.

He’s asking his people to ADAPT. NOW.

You might think you’re safe. You might tell yourself your job is too nuanced, too human, too... special. You know the type—the consultant who talks more than he listens, the lawyer surrounded by dusty tomes he hasn’t cracked open since law school, the senior strategist who spends more time refining his Notion workspace than his actual work. All of them still living under the illusion that AI is “not there yet”.

Spoiler: it’s already biting. Quietly. Efficiently. Without apology. Just go to Reddit and read the teary ‘I was just replaced by AI’ posts. It IS happening.

The memo reads like a pep talk written with a baseball bat. He tells his employees to scream in the mirror in anger if they need to. Then he tells them to study. Master AI. Build their own luck. Or get swept aside.

So yes, I could have written a few more paragraphs, thrown in a metaphor about boiling frogs or slow-burning wildfires. But honestly? Micha did it better this week. So I’m handing him the mic.

Read the memo below. Read it like your livelihood depends on it—because it just might.

And if you think this doesn’t apply to you… well, that’s exactly who it’s written for.

If for some strange reason you can’t read the pictures:

Hey team,

I’ve always believed in radical candor and despise those who sugar-coat reality to avoid stating the unpleasant truth. The very basis for radical candor is care. You care enough about your friends and colleagues to tell them the truth because you want them to be able to understand it, grow, and succeed.

So here is the unpleasant truth: AI is coming for your jobs. Heck, it’s coming for my job too. This is a wake-up call.

It does not matter if you are a programmer, designer, product manager, data scientist, lawyer, customer support rep, salesperson, or a finance person – AI is coming for you.

You must understand that what was once considered ‘easy tasks’ will no longer exist; what was considered ‘hard tasks’ will be the new easy, and what was considered ‘impossible tasks’ will be the new hard. If you do not become an exceptional talent at what you do, a master, you will face the need for a career change in a matter of months. I am not trying to scare you. I am not talking about your job at Fiverr. I am talking about your ability to stay in your profession in the industry.

Are we all doomed? Not all of us, but those who will not wake up and understand the new reality fast, are, unfortunately, doomed.

What can we do? First of all, take a moment and let this sink in. Drink a glass of water. Scream hard in front of the mirror if it helps you. Now relax. Panic hasn’t solved problems for anyone. Let’s talk about what would help you become an exceptional talent in your field:

Study, research, and master the latest AI solutions in your field. Try multiple solutions and figure out what gives you super-powers. By superpowers, I mean the ability to generate more outcomes per unit of time with better quality per delivery. Programmers: code (Cursor…). Customer support: tickets (Intercom, Fin, SentiSum…), Lawyers: contracts (Lexis+ AI, Legora…), etc.

Find the most knowledgeable people on our team who can help you become more familiar with the latest and greatest in AI.

Time is the most valuable asset we have — if you’re working like it’s 2024, you’re doing it wrong! You are expected and needed to do more, faster, and more efficiently now.

Become a prompt engineer. Google is dead. LLM and GenAI are the new basics, and if you’re not using them as experts, your value will decrease before you know what hit you.

Get involved in making the organization more efficient using AI tools and technologies. It does not make sense to hire more people before we learn how to do more with what we have.

Understand the company strategy well and contribute to helping it achieve its goals. Don’t wait to be invited to a meeting where we ask each participant for ideas – there will be no such meeting. Instead, pitch your ideas proactively.

Stop waiting for the world or your place of work to hand you opportunities to learn and grow—create those opportunities yourself. I vow to help anyone who wants to help themselves.

If you don’t like what I wrote; If you think I’m full of shit, or just an asshole who’s trying to scare you – be my guest and disregard this message. I love all of you and wish you nothing but good things, but I honestly don’t think that a promising professional future awaits you if you disregard reality.

If, on the other hand, you understand deep inside that I’m right and want all of us to be on the winning side of history, join me in a conversation about where we go from here as a company and as individual professionals. We have a magnificent company and a bright future ahead of us. We just need to wake up and understand that it won’t be pretty or easy. It will be hard and demanding, but damn well worth it.

This message is food for thought. I have asked Shelly to free up time on my calendar in the next few weeks so that those of you who wish to sit with me and discuss our future can do so. I look forward to seeing you.

Yours,

Micha

AI News

• A new study has raised concerns that LMArena — a leading AI model ranking site — may favor big tech models like OpenAI's and Google’s by allowing private tuning and uneven exposure, potentially distorting the leaderboard. Researchers say this boosts models through overfitting rather than true skill, raising questions about how we judge AI quality.

• Microsoft just released three new open-weight Phi models built for reasoning — with the largest outperforming much bigger competitors while still being small enough to run on everyday devices. The smallest, Phi-4-mini-reasoning, delivers strong math skills on phones, signaling progress toward powerful, locally run AI tools.

• Amazon launched Nova Premier, a powerful multimodal model designed more to train and improve its smaller models than dominate benchmarks itself. While it trails rivals in reasoning, it excels at managing complex agent workflows — part of Amazon’s strategy to scale capabilities efficiently across its entire AI stack.

• FutureHouse, backed by Eric Schmidt, just launched a suite of AI research agents aimed at supercharging scientific discovery — outperforming human experts at literature reviews and chemistry analysis. The four agents (Crow, Falcon, Owl, Phoenix) help researchers tackle the overwhelming volume of academic data, and are already available via web and API.

• Apple is reportedly teaming up with Anthropic to bring an AI coding assistant into Xcode, using Claude Sonnet to create a conversational “vibe-coding” tool for automating code writing, testing, and debugging. The move signals Apple’s willingness to partner with outside AI leaders to catch up after its rocky rollout of Apple Intelligence.

• Google released a sweeping energy policy roadmap and pledged to train 130,000 electricians to meet AI-era power needs — addressing looming shortages in both skilled labor and grid capacity. With AI driving explosive energy demand, infrastructure may now be just as critical to national competitiveness as chips and compute.

• OpenAI just reversed course on its for-profit pivot, announcing it will retain nonprofit control by restructuring its business arm as a public benefit corporation (PBC). The move comes amid pressure from former staff and legal battles, aiming to balance AGI goals with access to major funding — but raises questions about future investment terms.

• Over 250 tech CEOs — including from Microsoft, Adobe, Uber, and Khan Academy — signed a letter calling for mandatory AI and computer science education in U.S. high schools. The push coincides with a new executive order from President Trump and aims to ensure students are prepared for an AI-driven future.

• Nvidia open-sourced its new Parakeet V2 transcription model, which transcribes an hour of speech in just one second and tops industry benchmarks. With commercial-grade accuracy and a permissive license, the release lowers the barrier to building advanced speech tools across industries.

• Google quietly released Gemini 2.5 Pro I/O Edition, now ranked #1 on multiple AI leaderboards thanks to major improvements in coding, UI development, and even video understanding. It marks a major leap for “vibe coding” and shows Google continuing to deliver powerful updates without splashy launches.

• HeyGen’s new Avatar IV model creates hyper-realistic talking avatars from a single photo — now with hand gestures, emotion, and even side-angle generation. The upgrade supports various video formats and use cases, opening the door to more dynamic content beyond static talking-head clips.

• Lightricks unveiled LTXV-13B, a fast, open-source AI video model that generates high-quality footage 30x faster than competitors — and runs on consumer-grade GPUs. With pro features like camera motion control and keyframe editing, it's another sign that high-end video creation is being fully democratized.

• OpenAI just launched "OpenAI for Countries," a global initiative to help democracies build local AI infrastructure and tailor tools like ChatGPT to their specific cultural and civic needs. It ties into the $500B Stargate project and signals OpenAI’s intent to shape global tech influence under U.S.-led AI leadership.

• Figma unveiled a sweeping AI upgrade across its platform, adding natural language-to-code tools, one-click website publishing, smart vector editing, and AI-powered marketing asset generation. This positions Figma as a serious challenger to design giants like Canva, Adobe, and Webflow in the AI era.

• Mistral released Medium 3, a high-performing AI model with much lower costs than rivals, and launched its new enterprise platform with strong privacy controls and tool integrations. While Medium is closed-source, Mistral teased an open release of its more powerful Large model soon — a move that could shake up the open-weight space.

Quickfire News

• Anthropic launched Integrations, enabling Claude to connect with external Model Context Protocol (MCP) tools and expanding its capabilities with web support and new research features.

• NVIDIA pushed back against Anthropic’s AI chip export policy proposals, saying the U.S. should focus on fostering innovation rather than limiting its own global competitiveness.

• Google expanded AI Mode in Search to all Labs users in the U.S., and added new visual tools for shopping and local event planning.

• Suno rolled out version 4.5 of its music generation platform, offering more genres, improved prompt control, and the ability to generate songs up to 8 minutes long.

• Microsoft is reportedly integrating xAI’s Grok model into its Azure platform, amid speculation about growing tensions between Microsoft CEO Satya Nadella and OpenAI’s Sam Altman.

• Google released Little Language Lessons, a trio of AI experiments using Gemini to provide personalized, multilingual language learning tools.

• Google’s Gemini 2.5 Pro successfully played through Pokémon Blue, while an independent engineer streamed the run — a notable milestone following Claude’s still-unsuccessful gameplay attempts.

• Anthropic is reportedly conducting a $61.5B valuation tender offer, allowing employees and alumni to sell up to 20% of their shares, capped at $2M each.

• U.S. AI czar David Sacks predicted a 1,000,000x boost in AI capability within four years, citing explosive advances in compute, chip design, and model algorithms.

• Google is expanding Gemini access to users under 13, with strict safety guardrails and availability limited to supervised Family Link accounts.

• Google DeepMind’s Nikolay Savinov confirmed that 10M+ token context windows are coming soon, enabling superhuman coding applications.

• Zoom’s researchers introduced “Chain of Draft”, a prompting method that rivals Chain-of-Thought in accuracy while using just 7% of the token budget.

• AnySphere, the parent company of coding tool Cursor, raised $900M, bringing its valuation to nearly $9B amid growing demand for agentic developer tools.

• OpenAI CPO Kevin Weil said the company’s upcoming open-source model will reflect Democratic values and be one generation behind the frontier to prevent aiding adversaries like China.

• Pinterest launched new AI-driven visual search tools, enabling users to search using images instead of typed queries.

• Anthropic introduced “AI for Science,” a program that gives researchers free Claude API credits for use in fields such as genomics, agriculture, and drug development.

• The UAE announced mandatory AI education for all K–12 students, reinforcing its strategy to become a regional leader in artificial intelligence.

• OpenAI addressed GPT-4o sycophancy concerns, promising better evaluation standards, an alpha test phase, and increased testing transparency for personality and bias issues.

• OpenAI is finalizing a $3B acquisition of Windsurf (formerly Codeium), marking its largest acquisition yet as it deepens its investment in AI-powered developer tools.

• Google introduced AI Max, a Search-integrated advertising suite that helps advertisers automatically optimize reach and performance using generative AI.

• Microsoft is resisting OpenAI’s restructuring, seeking guarantees that its $13.75B stake is protected under the company’s new governance model.

• Elon Musk’s legal team criticized OpenAI’s PBC shift, calling it a “transparent dodge” that doesn’t address his ongoing lawsuit concerns.

• OURA debuted new AI features that help users track meals, monitor glucose levels, and receive personalized health advice via its smart ring.

• FutureHouse launched Finch in closed beta, an AI agent built to analyze and accelerate biological discovery using data-driven research tools.

• Apple may pivot Safari’s default search engine to an AI-powered system, with OpenAI, Anthropic, and Perplexity under consideration as alternatives to traditional search engines.

• Anthropic launched a web search API, letting developers enable Claude to search the internet and return citation-backed answers in real time.

• Google updated Gemini 2.0 Flash’s image model, improving text rendering, overall quality, and loosening previous content constraints.

• Netflix rolled out a new interface that adds an OpenAI-powered natural language search feature, letting users find content by describing it conversationally.

• LinkedIn introduced an AI-driven job search tool, allowing job seekers to use plain language to explore opportunities that align with their aspirations.

• Ace Studio released ACE-Step v1-3.5B, a new open-source music generation model capable of producing four-minute compositions in 20 seconds, including structural guidance.

Like newsletters?

Here are some newsletters our readers also enjoy:

Closing Thoughts

That’s it for us this week.

If you find any value from this newsletter, please pay it forward !

Thank you for being here !