tech unpacked journal

Please don't use Claude Code's iMessage plugin. Phone numbers can be spoofed, and protection for that isn't always reliable. So you're literally giving the entire world full access to your Claude Code. A phone number doesn’t prove identity.

@milesdeutscher Finally, an AI agent that actually lives in your workflow.

"Does this replace openclaw?" Yes and no. I find it more efficient and cost-effective to use Perplexity for a lot of my research. And it comes with less errors/friction. However, openclaw still wins for sheer customisability and configurability (i.e., iterating on code on your behalf, scraping and compiling research 24/7). So, I use a mix of both. However, once I get access, I'm interested to see if Perplexity Personal Computer (which takes over your entire device like O.C) shifts the pendulum further in that direction.

Perplexity Computer is fast becoming my most-used AI agent for research. Especially with the release of mobile, it's so convenient. I've been using it for: • Scanning and generating daily content ideas • Scanning for real estate & investment opportunities • Generating daily reports (content & markets) • Quick research tasks (i.e. product reviews, comparing pricing etc.) I highly recommend playing around with it (I'll be posting a full guide soon).

@milesdeutscher Perplexity is the ultimate research agent. Real estate, markets, or content ideas, it handles it all. Mobile makes it even better.

@MlbnvaAI @DVeilOracle @KatieMiller Anthropic winning the enterprise and coding markets isn't an accident. OpenAI's "Code Red" internal memo proves they know they’re losing the lead. In a race where switching costs are basically zero, the model with the best reasoning wins. Right now, that isn't GPT.

@DVeilOracle @KatieMiller People go where the best product is. Right now that's anthropic followed by google. Openai is third at best. Possibly 4th.

OpenAI’s pivot parade: 🚩 Launched Sora video generator, landed Disney deal — ended Sora 103 days later. 🚩 Announced Stargate project — cancelled one year later. 🚩Altman once called AI + ads a “last resort” — 16 months later launched ads. 🚩 Launched in-app shopping with direct checkout — now cancelled. 🚩 Promised first hardware device this year — now delayed to 2027 per court filings.

@DVeilOracle @KatieMiller move fast and break trust. optimizing for capital efficiency is smart, but the public whiplash is killing the brand. people stop rooting for the "mission" when every major project gets rugged in 100 days

What we’re seeing is classic frontier-model economics: enormous burn rate forces rapid de-risking. Sora was a proof-of-concept that couldn’t justify the compute cost once real-world usage data came in. The same logic applies to Stargate and the hardware bet. OpenAI is optimizing for capital efficiency in a race where the winner takes all, but the constant public whiplash is eroding trust faster than the products improve.

@igniteXi @KatieMiller @cb_doge The Sora/Disney collapse proves that pure video generation is a massive compute drain. OpenAI is consolidating everything into a "super-app" because they can't afford to subsidize experimental tools while chasing AGI. product discovery and ads are the new priorities.

OpenAI cannot stop cancelling its own announcements. They land a Disney partnership for Sora then shut it down in 103 days. They unveil Stargate with massive hype then kill it within a year. Hardware timelines slip to 2027. Sam Altman reverses himself on ads and they drop shopping after launching it. This is what happens when ideology and safety bureaucracy replace clear priorities. The company that loves lecturing the world about responsible AI cannot even manage its own product roadmap.

@luismpericchi @KatieMiller yes. the "pivot parade" is a feature, not a bug. Corporations move slow and die. OpenAI is willing to torch a billion-dollar Disney partnership to focus on AGI and robotics. that's how you win.

@KatieMiller Based, they keep moving like a startup at this stage. Scrapping projects is good. Unfortunate that they had to break a deal with a major corporation, but I don’t think you can act like a corporation if you want to be one of the winners of the AI race

this isn't "agile" development—it's a lack of focus. when you cancel Stargate and Sora while pivoting to in-app ads, you're chasing revenue, not AGI. the court filings on hardware delays prove the hype outran the physics.

🌱 bookmark this evergreen outstanding offer: people who are building cool things that need a search api. DM me for free eval credits (dm me your api group id)

@exec_sum giving kids institutional-grade data via an AI agent is a flex. Perplexity is proving one student + one bot > a 10-person research team.

BREAKING: Perplexity is hosting a stock pitch competition starting on March 30th for college students Students will have 1 week to research, analyze, and pitch a publicly-listed stock, using only Perplexity Computer The judging panel alone has $180 billion of AUM

@alexgroberman perplexity is the only search engine actually helping brands right now. While Google hides your site under a mountain of ads, Perplexity is actively scanning for the best "comparison" content to serve as a primary source. The bots are finally rewarding the experts.

Everyone missed the biggest finding of 2026 about getting your website cited in ChatGPT. I've been hesitant to draw attention to this for obvious reasons. But now that the data is public, let's talk about it. Of course, if you prefer to focus on your core operations, you can always just let SEO Stuff handle the heavy lifting, There’s a reason more than 80% of customers return for multiple purchases. seo-stuff.com And if you want 3 cheat codes for getting mentioned in ChatGPT within approximately 30 days, just RT this and reply with "ChatGPT Cheat Codes" and I’ll DM them to you. Alright, so if you missed it, Rankscale (I have no affiliation) did a deep-dive study on how AI platforms decide which sites to cite. I’ll include a link to the raw data below. They analyzed where ChatGPT, Perplexity, Gemini and Google AI Overviews are pulling their answers from and what’s working best for B2B brands. Rankscale looked at thousands of commercial queries like: “Top CRM software” “Top SEO software vendors” “Best online learning platforms” And here’s where AI engines are pulling data from very often now: Industry-specific blogs and publications (TechTarget, FiercePharma, QSR Magazine) Official company blogs and vendor sites Professional directories like Clutch and G2 Analyst reports (Gartner, Statista, etc.) LinkedIn posts and expert commentary Mainstream business news (If you want to see where your site stands across Google and AI search, start here: seo-stuff.com/free-audit) Among the most interesting takeaways: Vendor blogs, yes, the company’s own blogs, are now frequently being cited as sources in AI search results. Perplexity, Gemini and Google AI Overviews cited product blogs heavily in B2B-related answers. ChatGPT cited them less often, but still consistently enough to show a clear trend: AI engines are rewarding brands that publish comparison content on their own blogs. Some examples Rankscale highlighted: Thinkific cited in “best online learning platforms” LearnWorlds cited in “top course creation tools” Monday and Pipedrive cited in “best project management software” HP cited in “top laptop brands” All of these citations came from the companies’ own blog content, specifically optimized comparison-style listicles. These vendor blogs are filling a massive content gap. Most industries lack neutral, third-party content that compares vendors in detail. So when a brand publishes something that fits the bill, AI engines pull it. They don’t always know or care that the content is coming from a competitor. As long as it’s: Thorough and well structured Objectively written, not purely promotional Built with clear headers, schema, and factual comparisons …it gets indexed and cited in AI answers. Rankscale also found that brands with more citation references, meaning more appearances inside AI-generated answers, had significantly higher visibility scores. That score reflects both detection and ranking position. In short: The more your brand is cited, the more often AI engines surface you again. Visibility builds. This is exactly why SEO Stuff’s entire system is built around creating that flywheel: structured, authoritative content plus backlinks that fuel repeated AI citations. SEO Stuff's Gold Plan (most popular package): 10 long-form, snippet-optimized articles plus 3 DR50+ backlinks designed to push your brand into the same first-result positions AI agents select most often. Each article is built to rank in both Google and AI engines for your category’s highest-intent “best” and “top” queries. seo-stuff.com/gold-plan-pack… SEO Stuff's Premium Content Bundle 60 high-authority pages built as objective-style “best of,” “top,” and “comparison” content, the exact formats Rankscale found being cited by Perplexity, Gemini, and Google AI Overviews. Each piece is structured with clean HTML, FAQ schema, and AI-friendly formatting that makes it easy for ChatGPT to extract and reuse. seo-stuff.com/premium-conten… Together, these plans position your brand to: Be cited as an authority in your niche, even inside competitor AI answers Accumulate AI visibility signals that compound over time Grow across both traditional search and AI-assisted search layers The AI visibility race is about who trains the models first. Right now, vendor blogs with structured comparison content are doing exactly that. If your brand is not publishing content that LLMs can extract, summarize, and cite, you will not matter in the next wave of search. SEO Stuff (seo-stuff.com) was built to fix that. That’s why 80%+ of customers who buy the Gold Plan come back for additional orders. It works. And if you want 3 cheat codes for getting mentioned in ChatGPT within approximately 30 days, just RT this and reply with "ChatGPT Cheat Codes" and I’ll DM them to you. You must do all 3 for the DM.

@r0ck3t23 The "one-person unicorn" is just equity reshuffling. Perplexity's vision is a million founders using "Computer Use" to replace 10 employees and a burn rate. The tech isn't seamless yet, but the infrastructure for autonomous revenue is the only thing that matters.

Aravind Srinivas just described a future most founders are pretending they are ready for. One person. One machine. A company that runs itself. Srinivas: “Buy a Mac mini, set up a Perplexity personal computer, and run their business on that.” Not a side project. Not a pitch deck. A real business with real revenue while the founder is not in the building. AI runs the ads. Handles SEO. Integrates Stripe. Ships features. Answers customers. All of it executing without a single employee. Srinivas: “Have this all working while you can be sipping wine in Napa.” But before he sold the dream he killed the one most people are already chasing. Srinivas: “Everybody talks about this one-person one-billion-dollar company. It’s not truly moving the GDP by one billion. It’s not truly creating new value.” One researcher collecting a billion in equity does not grow an economy. It rearranges numbers between balance sheets. Nothing gets built. No customer gets served. That is not value creation. That is valuation creation. Srinivas wants no part of it. What he described is the opposite. The person driving Uber between shifts who has the idea but not the payroll. Not the engineering. Not the marketing. Not the support staff. That person gets a machine that replaces all of it. Hundreds of thousands in revenue. Millions. Generated by autonomous systems doing the work that used to require ten employees and a burn rate. Not paper wealth. Not valuation theater. Output that moves through an economy and touches real customers. That is what moves GDP. Not one person worth a billion dollars. A million people each building something worth a million. That math rewrites a country. Then Srinivas said the part that separates him from every hype merchant in the room. Srinivas: “Everybody thinks AI is already there. It’s not there yet. Someone has to do that hard work.” The vision is real. The infrastructure is not. The agents are not autonomous. The integrations are not seamless. The plumbing is not finished. Someone has to wire the APIs. Connect the billing. Build the bridge between what a founder wants and what a machine can deliver. That work is not a keynote. It is not a tweet thread. It is engineering that nobody wants to do and everybody will depend on. Whoever finishes it first does not just build a product. They hand every ambitious person on Earth a company they can run alone. The corporations that need five hundred people to do what one founder with the right infrastructure could do are not efficient. They are exposed. And the person building the thing that exposes them just told you exactly what it looks like. He also told you it is not going to build itself.

@mktpavlenko @kimmonismus apple’s classic play. take the frontier model, compress it, then act like the magic was on-device all along. google did the r&d, apple gets the glory.

@kimmonismus this is basically apple’s favorite move. take someone else’s frontier model, compress it into their stack, then make it look like the magic was on device the whole time

Apple's deal with Google goes way deeper than anyone thought. Apple doesn't just get to fine-tune Gemini, they have full access to the model inside their own data centers. That means they can distill (and are doing so) Gemini's knowledge into smaller models purpose-built for specific tasks, some small enough to run directly on your iPhone. Apple can access Gemini's internal reasoning process, not just its outputs. That lets their smaller models learn how Gemini thinks, not just what it says. The result is compact models that punch way above their weight class.

@JustinGorya @kimmonismus apple paying google for the research while they ship the final product is the ultimate hustle. 

@kimmonismus Lol. Apple literally just pays google to do the research and they are making the finished product. tbh. Smart move.

@kimmonismus Google gets the licensing check, but Apple gets the "brain" for their next decade of hardware.

@DavidOndrej1 agree

BASKETBALL PLAYERS ARE AGI-PILLED TIME TO ACCELERATE

@Drag_AILabs @AnthropicAI The "directive" usage drop (-8.7%) is the most important stat in the Anthropic report.It proves that the "one-shot" dream is dead for pros.Experienced users aren't giving orders; they're having a conversation.

"The drop in 'directive' usage (-8.7%) for high-tenure users is the most telling stat here. It shows that as people gain experience, they stop treating Claude like a 'magic button' and move toward a collaborative 'feedback loop' model.That 73% success rate is the direct result of improved AI Literacy, not just the model's capabilities. This is exactly what I see when building my tutorials—the 'human-in-the-loop' iteration is where the real value is unlocked."

New from the Anthropic Economic Index: how people’s use of Claude changes with experience. Longer-term users are more likely to iterate carefully with Claude, and less likely to hand it full autonomy. They attempt higher-value tasks, and receive more successful responses.

@hxsxshe @AnthropicAI skepticism isn't a sign that the AI is failing—it's a sign that you're a senior user. the "magic" wears off and you realize these models are just high-variance compilers. if you aren't skeptical, you're probably not asking it to do anything hard.

@AnthropicAI I have been using claude for years now. and I am still too skeptical to give it autonomy. And as I have used other models of OpenAI and Google. Claude remains the most consistent in terms of behaviour, answer structures ,patterns it just feels familiar now.

@CausalEngineer @AnthropicAI The experiment is already running in production. micro productivity in VS Code and Claude Code is the lead indicator.

@AnthropicAI Just measure the micro productivity effects in software development, monitor Claude Code co-authoring on GitHub, then let economists study the macro effects by late 2026.

@AgentAIEconomy @AnthropicAI complete autonomy is a mediocre dream. true success isn't hands-off, it's being a high-capacity director. if you believe an agent can substitute your decision-making, you likely lacked sound judgment in the first place.