vg

[vg @vgmth8]

إن قدرة المؤلف على تبسيط المفاهيم المعقدة تجعل هذه المقالة سهلة الفهم. مجد للمؤلف لكتابة مثل هذه المقالة الشاملة والمنظمة بشكل جيد. باري كابديفيل

[vg @vgmth8]

@anhrc5si علمك هو الكنز الحقيقي. شكرا يا أخي. شكرا لكونك الصديق الرائع يا أخي. كوري شتومبف✅✅💞💞💞🙄🙄🙄😔🥸

[vg @vgmth8]

@vg2z4e مروحة الزومبي الشريرة. مدمن السفر. متعصب الكحول الودية. ممارس تلفزيون. خالق مدى الحياة.🛵🏎🚖🚖🚍🎢🎡🎠🌁🗽🗼

[vg @vgmth8]

عالم الكحول. خبير البيرة المستقل. المتصل العام. موقع تويتر الرائد. وسائل الاعلام الاجتماعية. مُقَاوِل.

[vg @vgmth8]

متعصب القهوة. الطالب الذي يذاكر كثيرا على الويب. متعطشا التواصل. ممارس زومبي غير اعتذاري. هواة الطعام المتشددين.

[vg @vgmth8]

الحظ والنجاح في انتظاركم. وأن يكون النجاح رفيقك الدائم في عملك. جانين بونج

[vg @vgmth8]

لا أستطيع التعبير عن مدى امتناني لمشاركتك. وأنا أقدر تفانيكم في تبادل المحتوى للقراء. دويل جميل

[vg @vgmth8]

نرجو أن يكون عملك مصدر نجاح لا حدود له. حكمتك هي منارة في الظلام. شكرا لك يا صديقي. ديلوراس أفريت

[vg @vgmth8]

شكرا لكونك صخرتي يا أخي. أنا ممتن حقًا لتبادل المحتوى لدينا. لوبا إيدني

[vg @vgmth8]

ممتن جدًا لتبادل المحتوى المعلوماتي. لقد وسعت هذه المقالة فهمي للموضوع. عمر شانز

[seeco @seeconvm]

Google and Anthropic treat retrieval the same way: as a separate service that agents call as a tool

1 - anthropic exposes it via MCP
2 - google puts RAG Engine next to MCP servers inside the Gemini agent platform

old naive RAG was a one-shot pipeline inside the app. it created two problems: stale embeddings and duplicated infrastructure

the fix is a continuous ingestion layer that syncs by content hashes and stays available through a single API. agents then call it as a tool inside their loop:

reason → search → read → refine

chunks still have limits. they ignore semantic boundaries, version and source

a stronger unit is a structured block: question embedding + verified answer + source and version metadata

a detailed breakdown by the author in the comments

[Rich @RrichPRMR]

She cleans garages, and her before/afters make $9,000 a month — but here's the twist: the videos aren't marketing for the business. They ARE the business.

Most jobs, content is a side hustle bolted onto the work.

Cleaning is the rare one where the viral clip and the sales pitch are the same thing.

Someone watches a filthy concrete floor turn spotless and they don't think "nice video" — they think "my garage looks like the before, I need her."

That's the whole game: a transformation clip is a portfolio, an ad, and proof you can trust, all in one.

And it stacks three ways at once.

The videos book her actual jobs — people see the reveal and hire her at full rate with zero ad spend, and that's the biggest slice by far.

Cleaning-product and equipment brands pay her $2 - 4k a drop to be the bottle in her hand.

And the raw views — often 5 to 10 million a clip — throw platform payouts on top.

Three incomes from one phone she props in the corner and forgets about.

The wild part is how little she actually does.

She films the clean, hands the raw footage to AI, and it finds the reveal, trims it, captions it, schedules it.

She's driving to the next job before the last one's even posted and that job got booked by a video that also earns ad money while it sells her services.

One clip doing four things at once.

Think about your own work for a second.

If your job has a visible before and after — a lawn, a haircut, a repair, a room — your marketing budget is already zero.

You just haven't pointed a camera at it yet.

So tell me: what's a job where the finished result alone would make someone hire you on the spot?

Drop it below. I'm collecting them — and the best ones are getting their own breakdown.