Articles

Browse all AI articles. Use filters to narrow down results.

Showing 23641 to 23660 of 30623 entries.

Most Recent Articles

Product Hunt
startupproduct huntagentagents

DCP

By Iftakhar Rahmany

Discussion | Link

Ars Technica
aiai chatbotsai data centersai modelscolossuselon muskgrokorbital data centersspacexspacex ipoxaixai groktechars technicaopenaianthropicvision

SpaceX IPO filing pitches orbital data centers as Grok lags rival AI services.

Techmeme
techtechmeme

Rachel Metz / Bloomberg: Source: Cursor reached $3B in annualized revenue in late April and now has 3,000+ customers paying at least $100K each for its software on an annualized basis  —  Cursor's annualized revenue hit $3 billion in late April, according to a person familiar with the matter, indicating growing demand …

Techmeme
techtechmeme

Samantha Subin / CNBC: Spotify closes up 13% after announcing new features and 2030 guidance, forecasting a compound annual growth rate in the mid-teens and gross margins of 35%-40%  —  Spotify shares jumped 15% on Thursday after the music streaming platform laid out guidance for 2030 and reached an artificial intelligence deal …

Dark Reading
securitydark readingagent

How CISOs Should Prep for Agentic-Ready AI BOMs

By Ericka Chickowski, Contributing Writer

Finding ways to document both component and execution attributes for AI bill of materials (AI BOM).

Techmeme
techtechmeme

Brody Ford / Bloomberg: Zoom reports Q1 revenue up 5.5% YoY to $1.24B, vs. $1.22B est., AI Companion paid users up 184%, forecasts FY 2027 revenue above est.; ZM jumps 8%+ after hours  —  Zoom Communications Inc. projected stronger-than-anticipated sales growth, suggesting that customers are adopting its expanded suite of office products.

Techmeme
techtechmeme

Zaheer Kachwala / Reuters: Take-Two reports Q4 bookings flat YoY at $1.58B, forecasts FY 2027 bookings below est., reiterates GTA VI's November 19 launch date; TTWO jumps 6%+ after hours  —  Take-Two Interactive (TTWO.O) forecast annual bookings below Wall Street expectations on Thursday, but reiterated …

MIT Technology Review
artificial intelligenceroundtablessubscriber-only storiestechmit technology reviewmetallmvisionpolicy

Roundtables: Can AI Learn to Understand the World?

By MIT Technology Review

Listen to the session or watch below AI companies want to build systems that understand the external world and overcome the limitations of LLMs. Recent developments have brought world models to the forefront of the AI discussion. Watch a conversation with editor in chief Mat Honan, senior AI editor Will Douglas Heaven, and AI reporter…

Techmeme
techtechmeme

Jordan Novet / CNBC: Workday reports Q1 revenue up 13% YoY to $2.54B vs. $2.52B est., and lifts its full-year forecast, saying its AI strategy is working; WDAY jumps 9%+ after hours  —  Workday shares surged as much as 14% in extended trading on Thursday after the finance and human resources software maker reported results …

Techmeme
techtechmememeta

Ryan Autullo / Bloomberg Law: Texas AG Ken Paxton sues Meta, accusing WhatsApp of marketing its services as secure but failing to “deliver on those promises” by accessing encrypted messages  —  WhatsApp is able to access user's encrypted messages, Texas said in a lawsuit that accuses Meta founder Mark Zuckerberg …

Product Hunt
startupproduct hunt

buildpipe

By Raiyan Yahya

Discussion | Link

Techmeme
techtechmememeta

Erin Mulvaney / Wall Street Journal: Meta joins TikTok, Snap, and YouTube in settling with a Kentucky school district to avoid a trial over claims the platforms were designed to addict kids  —  Meta, TikTok, Snap and YouTube reached deal to avoid first of more than 1,200 consolidated lawsuits by school districts

The Conversation AI
aithe conversation aisafety

Would you trust AI to help you find ‘the one’? Dating apps are betting it can

By Maddison Sideris, Associate Teaching Fellow, Sociology, Deakin University

In the relatively short time they’ve been around, online dating apps have transformed how we hook up, date and fall in love. These apps have long been characterised by “swipe logic”: users are shown profiles of others and must swipe left or right to decide whether they like them or not. When Match Group first introduced this feature through Tinder in 2012, it fundamentally changed the dating scene. Now, another transformation may be on the horizon, as Bumble has announced plans to shift to AI-powered online dating and retire the swipe. The move raises concerns about the future of intimacy and dating. A Bee to find you love Chief executive Whitney Wolfe Herd recently announced that, at some point later this year, Bumble will be introducing an AI powered dating assistant named “Bee” which is expected to replace the swipe format. Bee will first chat to users to “get to know them”, before suggesting potential matches, and thoughtful date ideas. Bumble is the first mainstream dating app to overhaul its design to centre AI. But its not the first app to use it. Various dating apps have rolled out AI features over the past few years. Hinge, for instance, has incorporated generative AI tools to help users with conversation starters, and to give them feedback on their profiles, Bumble itself introduced an AI detective feature in 2024 to help boost user safety by identifying fake and scam profiles. The decline of dating apps Dating apps have become a key part of modern dating. They provide access to larger dating pools, especially for those experiencing “thin” dating markets due to their age, sexuality or location. Despite these benefits, however, research suggests individuals using these apps become increasingly burnt out and frustrated over time. Most major apps are now facing declining numbers in users and downloads – a shift characterised by swiping fatigue, safety concerns, and experiences of online racism and sexism. When it launched in 2014, Bumble was branded as the “feminist Tinder” because women had control over initiating the conversation. But it has faced various criticisms since then. For example, research published in 2018 argued the app was optimised to work best for straight, white, cisgender women – meaning they were likely to have smoother or more positive experiences than others. The app’s “women message first” feature also reinforces the traditional idea that there are two genders (men and women) and each one should play a certain role in relationships. The commodification of love Academics have been writing about the commodification of intimacy and dating since the late 1990s. Sociologist Zygmunt Bauman theorised we are living in a era of “liquid love”, in which we are “shopping for love”, and looking for shorter, more flexible relations. Another scholar, Eva Illouz, argued in her 2011 book Why Love Hurts that modern dating is shaped by an abundance of choice, and the pressure to compare and evaluate potential partners can make commitment feel more uncertain. Dating apps intensify these dynamics. They “platform” intimacy through a market-like system, where profiles (rather than people) are browsed, evaluated and selected. In this environment, it’s common to experience ghosting, stalled conversations, and ambiguity over what the other person wants. AI matchmaking might reduce the uncertainties of modern dating, by narrowing choices and recommending more “compatible” matches. Or, it might just shift the source of the uncertainty. Users will still have to navigate concerns around chemistry, intention and commitment – and now they’ll also have to decide if they trust the algorithm. They may not want to use AI at all, but will have no choice unless there’s a way to opt out. AI also introduces potential for bias in how users are “assessed” and matched. Although we don’t know exactly how Bee will work, a plethora of research has shown AI can perpetuate inequalities and lead to discrimination in various contexts. As American social scientist Safiya Umoja explains in her book Algorithms of Oppression, algorithms are not neutral systems; they can reflect and amplify existing social inequalities. On a dating app, this could look like reinforcing gendered dynamics and heteronormativity in who is shown to whom, reproducing racial biases already documented in online dating, and privileging certain lifestyles and markers of class in determining “compatibility”. Not to mention that if the algorithm is effective in finding matches “like you”, it could be closing you off from a variety of unique people and meaningful experiences. Is AI dating the way forward? In a society in which loneliness is on the rise, and love is in (apparent) crisis, do people want to use AI dating apps? It seems unlikely. Research indicates individuals are increasingly looking for more “organic connections”, and we have seen a boom in in-person dating events, from run clubs to social nights and speed dating. Outsourcing matchmaking to AI risks further reducing love and intimacy – a universal human experience – into something to be quantified and calculated, stripping it of its humanness. At the same time, we should remember people have a long history of getting external help to find love, whether that’s through paying a local matchmaker, posting a newspaper ad, getting advice from an agony aunt, or going on a dating show. AI could be the next iteration of how we manage the emotional toll of dating. So the question remains: would you trust AI to find you a partner? Maddison Sideris does not work for, consult, own shares in or receive funding from any company or organisation that would benefit from this article, and has disclosed no relevant affiliations beyond their academic appointment.