Author HubMaximilian Schreiner

Max is the managing editor of THE DECODER, bringing his background in philosophy to explore questions of consciousness and whether machines truly think or just pretend to.
AI research

OpenAI outperforms humans and Google at the world's top collegiate programming contest

AI in practice

New data from OpenAI and Anthropic show how people actually use ChatGPT and Claude

AI in practice

OpenAI launches GPT-5 as a unified system with adaptive reasoning for complex tasks

Ad
AI in practice
Short

OpenAI hit $1 billion in monthly revenue for the first time in July, CFO Sarah Friar told CNBC. The company expects to triple its annual revenue to $12.7 billion in 2025, according to earlier statements. By June, OpenAI had already reached an annualized revenue run rate of $10 billion. Most of this growth comes from paid ChatGPT subscriptions, especially since the launch of the new GPT-5 model. Despite some early criticism of the new model, Friar says OpenAI is seeing rising numbers of Plus and Pro subscribers. The surge in demand is putting pressure on technical resources - the company’s need for computing power now exceeds what’s available. Last week, CEO Sam Altman said OpenAI plans to invest in new data centers, with spending expected to run into the trillions.

AI in practice
Short

Gartner predicts that by 2028, one in four job applicant profiles will be fake.

According to a Gartner survey of 3,000 jobseekers, 6 percent admitted to cheating during interviews - either by posing as someone else or having someone impersonate them.

Gartner expects the rise of AI-powered tools like deepfakes, voice synthesis, and chatbots to make it even easier for fraudsters to fake skills or identities, all in pursuit of remote jobs and salaries. The Wall Street Journal reports that companies like Cisco and Google are already seeing these tactics and are bringing back in-person interviews to verify candidates and prevent abuse.

WSJ
Ad
Ad
AI in practice
Short

Google has started analyzing chat history by default in Gemini to personalize responses.

The company says the feature is switched on automatically so Gemini can deliver more relevant answers. Users who don't want their chats used for personalization have to manually disable the option.

With personal context enabled, Gemini saves key details and user preferences from past conversations and draws on them in future chats. The aim is to make interactions feel more like talking to someone who already understands the context.

This move puts Google on par with Anthropic and OpenAI, which both offer similar memory features.

Image: Google
Ad
Ad
AI and society
Short

Perplexity goes for another PR stunt: the AI startup offered Google $34.5 billion for its Chrome web browser, even though a sale isn't actually on the table. The company previously made public overtures toward TikTok in a similar bid for attention. With these stunts, Perplexity is trying to position itself alongside major tech brands, but it faces its own challenges, including multiple copyright lawsuits and mounting competition. The timing of the Chrome offer is notable, coming as a U.S. court weighs potential antitrust action against Google. According to the Wall Street Journal, Perplexity itself is valued at around $18 billion - about half the amount it offered for Chrome. Still, investors are reportedly willing to back the full amount. Perplexity says it would continue to support Chromium and keep Google as the default search engine. Most observers see the move as a play for attention, as Perplexity looks to stay relevant against rivals like OpenAI, xAI, and Anthropic.

WSJ
AI and society
Short

OpenAI is reportedly planning an investment in Merge Labs, a startup developing brain-computer interfaces that would put it in direct competition with Elon Musk's company Neuralink, according to the Financial Times. Merge Labs will be co-founded by Sam Altman, though he is not expected to take on an operational role. The company's goal is to use recent advances in AI to create more powerful brain implants. Merge is seeking to raise $250 million at a valuation of $850 million, the FT reports. Musk and Altman have been seen as rivals since Musk left OpenAI in 2018. Neuralink, which was founded in 2016, has so far raised $650 million from investors including Sequoia Capital.

FT
Ad
Ad
AI in practice
Short

Anthropic's Claude can now remember previous conversations with users and build on them over time. The new feature works much like ChatGPT's memory function, but puts a stronger emphasis on continuing and using information from earlier chats. According to Anthropic, the feature is now live for users on the Enterprise, Team, and Max plans, with support for other plans coming soon.

Google News

Bank details

IBAN: DE88 2507 0070 0053 0014 00
BIC: DEUTDE2HXXX
Account holder: Deep Content GmbH
Purpose: Support THE DECODER
AI and society

The long-predicted deepfake dystopia has arrived with Sora 2

AI in practice

Anthropic claims to lower the entry barrier for advanced AI models with Claude Haiku 4.5

AI research

OpenAI says GPT-5 shows 30 percent less political bias than previous models

Google News