AI assistants are becoming a new gateway to online services, and businesses that make their products easy for agents to use could gain a major advantage. A car-rental search illustrates the stakes: an agent gathered quotes from five of eight companies, while three blocked automated access. Rather than navigate those companies’ websites manually, the customer abandoned them. The shift is already visible in professional software, where AI-native companies such as Attio and Linear are winning users by treating agents as core users instead of simply adding chat features to older products. As major technology firms develop personal assistants, the companies controlling these gateways could shape how people find and buy things, much as browser, search, and operating-system makers have done.
Interfaces will still matter for visualizing complex information and for tasks people want to handle themselves, such as personal communication. But businesses’ own chatbots often lack the personal context that customers’ assistants already possess, forcing people to repeat details like travel plans, allergies, or addresses. Companies therefore need to distinguish legitimate purchasing agents from malicious bots, while making their services accessible through emerging standards such as Universal Commerce Protocol and Personal Agent Protocol. Walmart’s support for shopping through personal assistants and Shopify’s agent-ready storefronts show that this transition is underway. The argument is that agent access will soon be a basic requirement: customers will follow the easiest route to a purchase, and businesses that block their agents risk losing the sale.
Summary generated by AI from the linked article. hn.today is not affiliated with Hacker News or Y Combinator.
Researchers tested 16 LLMs on all 55 TIS-100 puzzles, giving each one attempt and time to refine its code. The models found 133 new performance tradeoffs across 18 puzzles and set three outright cycle records, though human players improved on most solutions within days. (tis-100.wincon.dev)
Australia is considering a new criminal offence for AI companies that fail to take reasonable steps to prevent harm caused by their models. The government is also weighing an AI safety watchdog with power to impose significant fines, following a hack of a Medicare statistics portal involving OpenAI. (abc.net.au)
AI is increasingly becoming a gateway through which people find information. Audiences still resist having AI produce their news. (niemanlab.org)
Nvidia holds a central position in the AI industry, making disruption to the company potentially catastrophic for the wider ecosystem. The race for advanced AI is also intensifying economic and geopolitical conflict, while concentrated corporate power and infrastructure face growing political resistance. (france24.com)
LLM inference can require reading about 26 GB of model weights to generate each token from a 13-billion-parameter model. The explanation shows how model size drives memory bandwidth demands during text generation. (youtube.com)
Deletion Tuning matches full fine-tuning across three LLM post-training methods by zeroing fewer than 0.05% of model weights. Its compact masks also preserve prior-task accuracy better than full fine-tuning, LoRA, and random sparse tuning. (arxiv.org)
Today's best Hacker News stories, summarized and screenshotted, one email a day.