6 Likes
6 Likes
8 Likes
6 Likes
OpenAI is said to be almost ready to unleash its own web browser, which could be out in the wild within weeks. According to Reuters sources, the company is aiming to more deeply integrate its services into usersâ work and personal lives, and the browser is part of that strategy (as is its push into hardware). Naturally, the browser is slated to have a ChatGPT-style chatbot baked in.
OpenAI is reportedly looking to use the browser to capture more user data â a strategy that has worked out to Googleâs benefit with Chrome. The browser is also expected to have agentic AI features such as Operator, which are billed as tools that can carry out actions (such as booking reservations) on a userâs behalf. Having direct access to information like web browsing data may make it easier for OpenAI to pull that off. âŚ
No.
Just: no.
Do not.
12 Likes
Not a fucking chance
8 Likes
9 Likes
Yes, please, take everything I hate and put it all in one place. They should get rid of it for competing browsers too just in case. I canât wait to not switch and have it all unavailable.
10 Likes
I mean, Iâd give it better than even odds that the reason they think they can get it out so fast is that theyâve forked Chromium and theyâre getting ChatGPT and vibecoders to make all the changes they think they need (and they asked ChatGPT what changes theyâre going to need), and it will download about half a web page before it explodes, and the codebase will be so utterly full of garbage that it wonât be fixable.
11 Likes
9 Likes
13 Likes
Perplexity rips another page from the Google playbook with its own browser, Comet
[âŚ]
The thing is, none of this really requires a dedicated browser. OpenAIâs Operator provides a way to automate browsing sessions and interact with the web either programmatically or through a chat interface. Anthropicâs Claude is capable of browser use.
But owning the browser provides access to a lot of downstream activity and data, and offers opportunities to strike deals with other vendors to make their services interoperate more effectively with AI agents.
Thus, thereâs a lot of AI-browser integration as browser makers work to keep control over how AI gets deployed and used. Google offers Gemini in Chrome. Opera is testing Opera Neon. Microsoft has installed Copilot in Edge. Mozilla has enabled a choice of chatbots in the Firefox sidebar. Brave has a built-in AI bot called Leo. Apple, sued for overpromising Siriâs capabilities, is reportedly looking into AI-augmented search in Safari.
Among commercial browser vendors, only Vivaldi seems willing to hold off on integrating AI with the browser.
[âŚ]
10 Likes
BTW
Whereâs Your Ed At:
Better Offline:
4 Likes
8 Likes
10 Likes
8 Likes
5 Likes
Dorsey, who also co-founded X (then Twitter), took to the social-media platform to announce Bitchat as a âdecentralized, peer-to-peerâ messaging app that âaddresses the need for resilient, private communication.â He called it his âweekend projectâ to learn about Bluetooth mesh networks, message encryption models, and more. Because it uses Bluetooth to function, rather than relying on the internet, the app is meant to be âresilient to network outages and censorship,â according to its technical whitepaper on GitHub.
Shortly after Bitchatâs unveiling, Radocea, whose company builds secure Wi-Fi routers, wrote a blog post that pointed out a key vulnerability in the app. In the post, he called âidentityâ Bitchatâs âmost glaring issue.â He pointed out shortcomings in user authentication and verification. Attackers can present other usersâ identity keys to imitate their trusted âfavoriteâ contacts. He said the shortcomings are âthe hallmarks of vibe code (in)security,â or using agentic generative coding.
âBecause it was built with GenAI, I think to [Dorsey], when he was writing it, it looked like it was doing the cryptography it should have done, when that wasnât the case at all,â Radocea says. âAnybody could show up pretending to be anybody else, because the gen AI code that he put together just does not do cryptography for the identity verification between the two parties, so you donât really know who youâre talking to in the chat.â
But bitchatâs most glaring issue is identity. Thereâs essentially no trust/auth built in today. So I would not really think about this as a secure messenger. The protocol has an identity key system, but itâs only decorative as implemented and has misleading security claims. The 32-byte public key gets shuffled around with ephemeral key pairs as an opaque blob. The user verification is unfortunately disconnected from any trust and authentication. These are the hallmarks of vibe code (in)security.
8 Likes
Artificial intelligence promises a future of unprecedented productivity and wealth
Meanwhile, Bernie has his eyes on the prize hereâŚ
In a conversation with Gizmodo, the Vermont senator, who revealed he had just spoken with one of the worldâs leading AI experts, laid out his fears that the technology will be used to suppress wages, break unions, and further enrich the billionaire class. He also shared his concerns about AIâs impact on our collective mental health and discussed the âdoomsday scenarioâ that has some top minds in the industry worried that humanity could lose control of its own creation.
9 Likes