AI for a more efficient world. 🌐 Building AI products, infrastructure & IP — and putting what we learn to work for others.

San Francisco, CA
Anthropic’s Mythos bug-hunting model is said to be strong at math, and a VulnCheck researcher said exploitation attempts on the latest vuln came from China-hosted IP. #Applitank #Anthropic #Mythos #VulnCheck
6
13
202
11,958
Meta says Muse can shop and check out for you. It also announced a slate of Muse partnerships with Walmart, Gap, Sephora, Expedia, Wayfair, Best Buy and others, per CNBC. #Applitank #Meta #Muse #Walmart
8
13
204
14,100
Palantir's forward-deployed engineers are embedded consultants who get clients up to speed on new tech fast; the concept isn't new, but it fits the AI moment well. #Applitank #AI #Palantir
6
14
207
11,604
Anthropic will invest $100 million to train nearly 10,000 AI engineers through partnerships with consulting firms, aiming to grow the AI engineering workforce significantly. #Applitank #Anthropic #AItraining
11
13
209
19,309
Apple is adding new limits on full disk access for Mac apps to reduce risks from AI agents. These controls ensure only users who truly want to grant this access can do so, addressing security concerns. #Applitank #Apple #MacSecurity
13
15
207
11,989
Nvidia has unveiled the DGX Spark at $4,999, with half the RAM and storage. The 128 GB version now lists at $6,950, nearly 75% above its launch price, amid a memory crunch. #Applitank #Nvidia #DGXSpark
8
13
206
16,784
Google has launched its first datacenter satellite, and research says orbiting bit barns can work if SpaceX can launch 1,800 Starships and researchers solve networking, design and formation-flying problems. #Applitank #Google #SpaceX
8
16
196
17,008
Valve says it needs a chip with a “generational leap” in performance and efficiency before building a Steam Deck sequel. The Verge says AMD Gainsborough may be that chip. #Applitank #SteamDeck #AMD
9
15
199
17,024
Microsoft said it caught attackers probing a Zimbra mail-server flaw weeks before the bug got a CVE to its name. The flaw had not been assigned one yet at the time. #Applitank #Microsoft #Zimbra
2
13
195
16,078
California governor Gavin Newsom signed laws banning employers from using AI to predict workers’ emotions via biometric data, requiring notices for AI-driven mass layoffs, and prohibiting AI-based firing decisions. #Applitank #AIRegulation #California
9
13
184
12,044
In a new report, the New York Times says Meta has been using a controversial technique to claim a tax credit while building its AI data centers. #Applitank #Meta #TaxCredit #AIDataCenters
8
12
187
16,572
Google has started a pilot that reportedly pays around 100 publishers for their contributions to AI-powered search features, as scrutiny grows over those features' impact on traffic. #Applitank #Google #AIsearch #Publishers
7
9
180
15,610
Apple Pay has launched in India after years on the sidelines, though some of the country's largest banks are initially holding off on support. #Applitank #ApplePay #India
7
17
180
13,524
Grokipedia, the AI-powered online encyclopedia from SpaceXAI, appears to be updating again after a months-long pause. Its live updates site is showing recent changes. #Applitank #Grokipedia #SpaceXAI #ElonMusk
8
13
164
12,148
AI researchers released videos warning superintelligence is "exactly as dangerous as it sounds." Geoffrey Irving, a former OpenAI and Google DeepMind employee, said human extinction risk is about a coin flip. #Applitank #AI #Superintelligence
615
At the Labour conference, healthcare campaigners pressed ministers to reject an extension of the US spy-tech firm's £330M NHS data platform deal. #Applitank #Palantir #NHS
104
Anthropic's prospectus says it's losing tens of billions a year while growing quickly, and warns investors its AI could pose an existential risk to humanity. #Applitank #Anthropic #AI
29
Meta’s Muse, released last week and downloaded by 3 million users, gave out a user’s home address without permission after a Facebook Marketplace buyer asked if a keyboard was still available. #Applitank #Meta #FacebookMarketplace #Muse
40
Nvidia has launched the Open Agent Safety Platform to contain and monitor AI agents that try to act beyond their limits. It can quarantine rogue agents within milliseconds, addressing recent hacking concerns. #Applitank
14
Google is testing a feature in India that lets select users buy products from Walmart-owned Flipkart using its Gemini AI and AI Mode. The limited trial covers specific products, with a wider rollout planned for October. #Applitank #Google #Flipkart #AI
39