Loading activity…
@mparakhin
Shopify CTO
Technology
Google DeepMind released Gemini 4 Argon, its flagship model, via Gemini app, API, Google Cloud, and Search AI Mode.
AI
A post sharing a conversation with Amplitude, Deloitte and Rakuten leaders says the model scored 90%+ on data visualization evaluations.
AI
Yann LeCun recalls Minsky's reaction to his experiment training a neural net to multiply two binary numbers.
AI
A user argues that pretraining on mediocre code leaves habits that post-training cannot entirely erase.
AI
One poster misses LAN parties; another recalls getting a first internship and falling in love with machine learning that year.
AI
A post quotes AI as a 'superhuman specialist' that's 'sub-human at running itself,' then speculates about what that gap could lead to.
AI
Perplexity claims leading Answer and Evidence retrieval on context-bench, while its earlier model leads two Document measures. The preview’s embeddings may be incompatible with future releases.
AI
A user quoted Pro outlining a technical approach, replying to someone who said they get excited about finding problems that need that level of mathematical horsepower.
AI
Perplexity says Fast Search returns 95% of search results in 230 milliseconds or less. It runs on Photon, which the company describes as its new Rust-based retrieval and ranking service.
AI
One post describes xAI as three years old—half Anthropic's age and a quarter of OpenAI's—and argues that the AI race is not over.
AI
One commenter sees a largely linear increase in quality as reasoning effort rises.
AI
TIME describes the list as 50 leaders shaping technology, artificial intelligence, cybersecurity and digital innovation.
AI
OpenAI says Sol and Luna’s API prices are 50% below GPT-5.6. A launch-day recap cites Anthropic’s claim that Opus 5.5 costs around 40% less than Opus 5 on typical workloads at default settings.
AI
One tester rates GPT-6 Pro as the best model yet for math, machine learning and brainstorming, but says it sometimes loses to Astra Max.
AI
A user says they kept the same setup and ran both models at Max effort. They interpret the result as a roughly tenfold increase in model intelligence over nearly six months.
AI
One user questions the fascination with command-line interfaces, calling Claude Code's sidepane emulation impressive but arguing for investment in the app instead.
AI
A user gives Applied Compute a strong endorsement while saying they have no relationship with the company.
AI
Creator Diogo Almeida claims Jev is 20–200x faster and 40–400x cheaper, with free output tokens. Investor DCVC says it led TypeSafe’s $40 million seed round.
AI
A reader argues Michael Lewis misjudged Google in 1999, using a passage from The New New Thing as a caution about predictions during the AI boom.
AI
Shopify CTO shares early tests of GPT-6 and Fable 5.1
