In this article, we have put together a list of the 20 most popular tech jobs in Asia in the past week. Got your resumes and job profiles ready? Check out the opportunities below, and apply away (:

If you’re hiring, you can post your jobs for free here.

This article summarizes First Round Capital’s podcast featuring former Twitter CEO Parag Agrawal.

Parag Agarwal, founder of Parallel and fomer CEO of Twitter / Source: Twitter

The web faces an existential risk of closing as data producers build paywalls against high-volume AI traffic. Parag Agrawal, ex-CEO of Twitter and co-founder of Parallel, argues this shift requires a rebuild of the web’s technical and economic infrastructure to serve its emerging primary user: AI agents.

This evolution away from human-centric design means the industry must create new product development frameworks and economic models. These changes are necessary for navigating a future where AI usage will dwarf human activity by orders of magnitude.

AIs will become the web’s primary user by orders of magnitude

The web is currently transitioning away from human-led interaction toward AI-driven processes. Agrawal argues that this evolution necessitates a fundamental overhaul of existing infrastructure to manage the unprecedented scale of non-human activity.

The core premise is a massive user-base shift
Agrawal states, “we believe that AI will use the web a thousand times, a million times, more than humans ever have. And as a result, the web will need to transform. In order to drive that transformation, we’ve been building the best tools that AIs can use to access content in the web.”

This change requires a new architectural starting point
Agrawal explains his thinking, “I started really living in this notion of, ‘All of that’s going to change completely.’ All of the infrastructure that we built, thought about, it’s going to look completely different. Every business model we thought about is going to look completely different. So, it really started as this science fiction of the primary consumer on the web is now going to be an agent.

The new web must be built for an entirely different user psychology

Human users operate within a narrow band of patience and specificity. They require sub-second responses and often provide vague queries. AIs, however, can be highly specific and are often indifferent to latency, which allows for a new class of “deep research” applications.

Human users are constrained by impatience and ambiguity
Agrawal notes, “humans, as I frame it, operate in a very narrow band. We have a second or two of patience. We under specify what we’re looking for. We will either implicitly just click on an app and expect the app to figure out what we want and show it to us, like Twitter, or we will type an incomplete set of keywords.”

AI users have fundamentally different needs and capabilities
Agrawal continues, “AIs can specify what they’re trying to solve… the problem space really expands. You no longer are stuck with producing a consistent format in your answer or a set quantity in your answer or a 1-second latency on your answer. Sometimes you’ll want things in 10 minutes and sometimes you’ll want 100 documents and sometimes you’ll want a one-word answer.”

The current web economy is misaligned with AI usage

The existing ad-supported model is not equipped for massive-scale AI traffic. This misalignment creates an economic incentive for data producers to put their content behind paywalls. This trend threatens to fragment and close the open web, creating an existential risk.

The open web is at risk of closing
Agrawal warns, “I think there is an existential risk on the web that the web might close up more and more unless we figure out and solve problems in a way that incentivizes this data to remain open. And it must be optimal for most people… that their goals [are better served] to be open rather than paywalled or closed.”

A new model must enable differential pricing at scale

To keep the web open, Agrawal looks to the advertising business for a new economic model. His idea involves “differential pricing,” where high-value commercial AI queries subsidize low-value personal usage.

The advertising model offers a precedent for differential pricing
Agrawal explains the analogy, “what ads did really well was differential pricing at scale and efficiency… they make money from a small fraction of users which allows them to be accessible to a wide, wide, wide swath of users… they lose money on a large number of users but that’s how the business works.”

This model can be applied to AI’s web consumption
Agrawal uses his interviewer’s firm, First Round Capital, to illustrate the concept: “if you consume someone’s content through some really expensive [AI] model and spend like US$4 in GPU to get some work done at First Round, versus my retired dad just reading the news… I’d like for you to subsidize that. So you shouldn’t have to pay the exact same for the same content.”

Product development is shifting from human needs to technical possibility

Pre-AI product development focused on the intersection of human needs and what could be built. In the AI era, human needs are often assumed to be vast. Agrawal believes the new focus is on what is technically possible and anticipating the rapid trajectory of model improvements.

• Old model: Identify a human need, then determine if technology can solve it. The core question was the intersection of user demand and technical feasibility.
• New model: Assume the need exists. Focus entirely on what works, what can work soon, and building a step ahead of the current capabilities of rapidly advancing AI systems.

System architecture must now embrace stochasticity and relaxed constraints

Classic web development optimized for deterministic systems that delivered results with sub-second latency for impatient humans. Building for AI agents allows for a new architecture that embraces stochastic, probabilistic systems and can relax constraints like speed to maximize quality.

• Systems are moving from deterministic to stochastic, where outputs are probabilistic.
• This requires a new way of building and communicating about products and their reliability.
• Constraints designed for humans, like the need for speed, can be relaxed for AIs.
• By relaxing latency requirements, systems can perform deeper, higher-quality work.

A counterintuitive design choice is to build for slowness
Agrawal explains, “we took the counter approach of saying our customer is an AI. And one of the things that led us down was we are going to design an API that is excruciatingly slow. Humans have no patience on the web… If it’s an agent doing something in the background… the latency of each operation doesn’t matter. And if you relax that constraint, you get to do more.”

This summary was generated with AI. A human editor reviews it before publishing.