Instinct

Instinct

2026-10-05 19:01:24

Instinct rolls out group chat AI agents for collaborative tasks with friends

Instinct, an AI agent company, has introduced a group chat feature that lets users work with friends through its AI agents on shared tasks. The company said the feature is designed for collaborative use cases such as planning trips, organizing carpools, and coordinating activities. Instinct also said individual account information remains separate, addressing how personal data is handled when multiple people use agents in the same conversation. Before a personal agent shares information or takes action, it must first obtain permission, according to the company. The update was reported by Techub News, which cited TechCrunch.

390

Instinct

2026-09-30 07:51:41

Instinct adds Selections feature, bringing expert picks into its personal agent

Personal agent assistant Instinct has launched Selections, a new feature that plugs recommendations from real-world experts into its agent product and then filters those options based on a user’s preferences. The company said it plans to work with local chefs, designers, architects, and travel guides, who will curate restaurants, travel routes, and products in their areas of expertise. Instinct then uses those lists to surface choices that match a user’s style, budget, or other stated needs. The rollout draws attention because Noah Shinn recently argued against an ad-driven model in his first public long-form interview. Shinn said personal agents may hold large amounts of user data, and letting brands pay to persuade users to buy things they did not originally want would create an obvious conflict of interest. Instinct’s current plan is to offer the product to users for free and charge commissions to merchants that receive orders through the agent. The company has not said whether merchants can pay to enter recommendation lists, and it has not disclosed whether participating experts are being paid. That leaves an unresolved question over whether future recommendations could be influenced by how much revenue Instinct can earn from a transaction.

870

Instinct

2026-09-29 06:57:55

Instinct founder says user growth of 10% to 11% a day is pushing compute demand close to doubling weekly

Instinct founder Noah Shinn said the personal AI agent startup has been seeing user growth of roughly 10% to 11% per day in recent weeks, a pace that has pushed its compute needs close to doubling every week. He said around 40% of his time is now spent deciding how much capacity the company should buy in advance, because compute procurement cannot keep up in real time with user growth. Shinn laid out the trade-off in simple terms: buying 2x current capacity would last about a week, while 5x would be filled in less than three weeks. Buying 10x, on the other hand, would require a large upfront cost. He added that many resources must be reserved months ahead, and emergency capacity can cost 3x to 4x more. Even if daily growth slows from 10% to a range of 5% to 8%, Shinn estimated that after three to four months the company could still need to prepare compute for a user base on the scale of 100 million. He said the wait time for new capacity to come online is roughly in the same range. In his view, scaling a personal agent product is harder than scaling consumer internet apps such as Instagram or Facebook because each added user continuously consumes significant model inference.

790

Instinct

2026-09-29 05:02:46

Instinct says 40% of users grant credit card access within three weeks as founder admits early version lacked firewalls

Instinct founder Noah Shinn said about 40% of users allow the product to use their credit cards after three weeks, letting its agent complete purchases on their behalf. He said the company tracks how long it takes before users first share sensitive information such as credit cards, account passwords, or other credentials as a measure of trust. Once a user connects at least one piece of sensitive information, retention is about 80%. Shinn also acknowledged that Instinct’s early version did not have the firewalls and active monitoring systems now in place, and that the product made mistakes. He said the company later rebuilt its security architecture instead of patching isolated issues one by one. Under the current setup, text, images, and other content are screened by an independent firewall before being handled by the agent, while a separate system monitors planned actions and parts of the model’s reasoning so they can be paused or blocked before execution. Shinn was also responding to a prior user report that Instinct had started processing a financial document the user said had never been uploaded. He said the incident was a hallucination in which the model fabricated information and then fabricated an explanation, not a user data leak. Instinct later launched a standalone hallucination detection system to check and intercept anomalies before the agent continues reasoning or calls tools.

710

Instinct

2026-09-29 04:55:29

Instinct says 10% daily user growth is pushing compute demand close to doubling every week

Instinct founder Noah Shinn said the personal AI agent startup has been growing users by roughly 10% to 11% a day in recent weeks, a pace that is driving compute demand to nearly double every week. He said about 40% of his time now goes into deciding how much capacity to buy in advance, because compute cannot simply be added at the same speed users arrive. Shinn laid out the trade-off in blunt terms: buying 2x current capacity may last only about a week, while 5x capacity could still be filled in less than three weeks. Buying 10x capacity, on the other hand, means taking on a large cost upfront. He added that many resources must be booked months ahead, and spot purchases can cost 3x to 4x more. Even if daily growth slows from 10% to a range of 5% to 8%, Shinn estimates that after three to four months of continued expansion, Instinct may still need to provision for a user base of 100 million. That timeline, he said, overlaps with the waiting period needed for new capacity to come online. He argued the scaling problem is harder than it is for traditional consumer internet apps, because each new personal agent user continues to consume heavy model inference.

570

Instinct

2026-09-29 04:07:16

Instinct eyes free AI agent model with revenue tied to transaction commissions

Instinct founder Noah Shinn outlined how the personal AI agent startup may eventually make money in his first public long-form interview. His stated goal is to keep Instinct free for users over the long run rather than charging subscription fees. Instead, if the agent helps users complete actions such as hotel bookings or shopping purchases, Instinct would take a small commission from the merchant that wins the order. Shinn said that remains a goal, not a firm commitment to permanent free access. He also drew a clear line against advertising. Shinn said personal agents will become more familiar with users over time and more capable of shaping their decisions. In that context, he argued, an ad-driven model could push an agent to persuade users to buy things they did not originally plan to purchase. He said he would rather have Instinct earn money only from transactions users were already intending to complete. The exact commission rate has not been set. Shinn referenced platform take rates used by Shopify, Amazon, and Apple. He added that about half of Instinct’s transaction volume currently comes from travel, and that some boutique hotels are willing to pay commissions of more than 30% for orders the platform brings in, though Instinct may not charge that much.

540

Artificial In

2026-09-28 11:36:47

Instinct raises $1 billion at a $10 billion valuation

ChainCatcher reported that artificial intelligence company Instinct has raised $1 billion at a $10 billion valuation. The brief did not provide additional details such as the financing round, participating investors, or how the proceeds will be used. Based on the information available, the disclosed facts are limited to the company name, the amount raised, and the valuation attached to the deal. The update was published by ChainCatcher on September 28, 2026.

540

Instinct

2026-09-21 03:03:39

Reverse engineering suggests Instinct stores long-term memory in Git-managed Markdown files

Supermemory founder Dhravya Shah said his reverse engineering of the personal agent assistant Instinct points to a long-term memory design built around Git-managed Markdown files rather than replaying entire chat logs into context. In Shah’s account, Instinct keeps an up-to-date user state, then uses a separate file-based store for more detailed records such as preferences, decisions, tasks, and person-specific facts. His black-box testing suggested that each conversation begins with a user state summary of about 4,250 tokens, covering background, operating preferences, and communication habits. Shah said he did not find signs of vector indexing or BM25 retrieval. Instead, detailed memory appears to be fetched with keyword search similar to grep, while a background process continuously reorganizes information: temporary details are grouped under tasks, experiences are distilled into long-term preferences, old facts are revised with new ones, and Git preserves version history. He added that the main agent appears to focus on reading memory, while the cleanup and merge work happens in backend workflows that he estimates run about once a day. According to Shah, the setup is close to Supermemory’s own view of how agents should manage memory, and he said the architecture can be largely replicated with Supermemory. He shared an example that used roughly 60 lines of code to plug user profiles, memory search, version history, and continuous updates back into an agent.

1050