AI assistants

AI assistants

2026-09-15 04:14:41

Assistant Benchmark launches real-world leaderboard for personal AI agents, with Muse on top for now

Assistant Benchmark, an independent evaluation project, has started comparing personal AI assistants through real-world tasks rather than abstract benchmarks. The test setup asks agents to complete practical assignments such as booking hotels, choosing restaurants, shopping, replying to emails, and connecting third-party services. Scores are based on whether those tasks are actually completed, and each scoring category is tied to a public fixed task that must include a real execution record before a score is awarded. At the current stage, Muse ranks first with an average score of 9.1 across seven completed categories. Instinct has been tested across 11 categories and holds an average of 8.4, while Grok Bot has completed seven categories with an average of 7.3. Muse posted perfect 10-point results in shopping, email, and third-party app connection tasks. The benchmark should be read as a live, evolving experience board rather than a final ranking. Muse’s 9.1 only reflects the average from seven tested categories, not a full result, and each category is still being measured with only one fixed task for now. As more follow-up testing is added, both the scores and the ranking may change.

550

Counter-posit

2026-09-11 09:03:12

Positioning Against Incumbents: Why Startups Need a Moat Big Companies Can’t Copy

A TechFlowPost translation of a Not Boring essay argues that the most useful early moat for startups facing dominant incumbents is not simply building a better product. It is "counter-positioning" — designing a business model that established players cannot follow without damaging their own economics. The piece walks through examples including Ramp, Base Power, Dell, Android, Amazon, Microsoft and Facebook to show how startups can exploit the constraints of larger rivals during the takeoff phase. The article also stresses that counter-positioning is temporary by nature. Hamilton Helmer, whose framework underpins the discussion, describes it as an incomplete source of power: enough to create a window, but not enough to secure lasting durability. To survive after that window opens, startups still need long-term defenses such as scale economies, switching costs, network effects, brand or process power. That framework is then applied to the AI assistant race, especially Instinct and Meta’s newly released Muse. The argument is blunt: being better is an advantage, but not a moat. If a company like Meta can copy the product without hurting its existing business, the startup has little time to turn product momentum into a real barrier.

410

Apple

2026-08-04 03:37:49

Apple’s new Siri may dominate on installed base with iOS 27, but ChatGPT still leads on conversation

Apple is expected to roll out a revamped Siri in September alongside iOS 27, instantly giving it the largest installed base of any AI chatbot because the assistant will ship preloaded on hundreds of millions of iPhones, including devices going back to the iPhone 15 Pro. But a hands-on test by Bloomberg reporter Natalie Lung found that scale does not automatically translate into stronger chat performance. According to the report, the new Siri’s edge comes from device awareness. It can read emails, messages, notes and files stored on the phone, and it can also understand what is currently shown on screen. That allows it to generate more tailored responses and help with items such as spreadsheets, documents or even crossword clues. The trade-off is setup time: Apple’s assistant may need hours or even days to re-index device content before it can retrieve older information reliably. The report also says Siri still lacks access to real-time data from third-party apps, while ChatGPT already offers broader external integrations. OpenAI, for its part, is pushing ahead on voice interaction and productivity tools. Bloomberg highlighted GPT-Live for more natural spoken exchanges and ChatGPT Work for integrations with Codex, Google Drive and desktop workflows. In short, the comparison is shaping up as distribution versus capability: Apple has the channel, while OpenAI still holds the lead in conversational fluency and workflow breadth.

1910

MoonPay

2026-07-31 16:55:51

MoonPay launches PayBox for crypto payments through AI assistants

MoonPay has introduced PayBox, a non-custodial payment vault that lets users carry out crypto transactions and online purchases through AI assistants including ChatGPT and Claude. The product supports Solana as well as Ethereum Virtual Machine-compatible networks, extending its reach across multiple blockchain environments. PayBox works through custom connectors and natural-language commands, allowing users to instruct AI tools to handle specific payment and transaction tasks. To control those permissions, users can set Passkey-based authorization and spending limits. The product also supports a range of use cases beyond basic transfers, including digital asset purchases, cross-chain bridging, and commercial payments such as booking flights and restaurant reservations. For security, private keys are stored using multi-party computation shard storage, according to the report cited by Techub News from crypto.news.

1960

ChatGPT

2026-07-23 10:35:15

ChatGPT Market Share Falls Below 50% as Gemini Gains and Claude Leads in Paid Conversion

Sensor Tower said ChatGPT's share fell to 46.4% by late May, while Gemini reached 27.7% and Claude 10.3%. In the US, Claude posted the highest paid subscription conversion rate at 13%.

470