Midjourney Enters Video AI, Sequoia Bets on Lawyer Bots

Midjourney released its first AI video generation model, V1, which transforms a single image into four five-second clips. The company prices each video prompt at eight times the cost of an image generation.

V1 is offered through subscription tiers starting at $10 a month, with Pro and Mega plans granting unlimited video generations in slower mode. Midjourney says it will revisit its video pricing in the next month and is already planning AI models for 3D rendering and real-time simulation.

The rollout puts Midjourney in direct competition with OpenAI, Runway, Adobe, and Google’s video tools. It also arrives one week after Disney and Universal sued the startup for alleged copyright infringement tied to its image models.

AI art residencies are springing up across Europe, North America, and Asia, putting generative tools directly into artists’ hands. French-American cultural body Villa Albertine has launched a dedicated AI track that will add four fellows a year to its 60-resident program.

Residencies at institutions like Mila, the Max Planck Institute, and the SETI Institute supply compute, collaborators, and exhibition pathways, with alumni works shown at MoMA and Centre Pompidou. Villa Albertine announced its initiative at a Paris AI summit attended by Culture Minister Rachida Dati, OpenAI applications chief Fidji Simo, and more than 500 industry participants.

Program organizers say the residencies avoid “choosing sides,” yet ethicists note the curated setting can soften public skepticism toward AI-generated art. Copyright lawsuits against Stability AI and Midjourney underscore that legal questions around authorship and compensation persist despite the cultural embrace.

Multiplier Holdings secured $27.5 million across seed and Series A rounds to expand its AI-driven acquisition model for accounting firms. The former Stripe executive–led company buys small service businesses and outfits them with its proprietary automation tools.

Lightspeed Venture Partners led the Series A while Ribbit Capital and SV Angel backed the earlier seed. Multiplier proved the concept by buying 12-person Citrine International Tax and more than doubling its profit margins after embedding AI.

VCs are flocking to AI-powered roll-ups, with General Catalyst, Elad Gil, Thrive, and Khosla Ventures backing similar plays. Lightspeed partner Justin Overdorff says the model only became workable once AI eliminated manual work, and he has already funded three more stealth companies using it.

Crosby exited stealth with a $5.8 million seed round led by Sequoia and Bain Capital Ventures. The firm offers contract-review services that combine in-house AI software with its own team of lawyers.

Crosby employs 19 people and has already reviewed over 1,000 contracts for startups such as Cursor, Clay, and UnifyGTM since a January soft launch. Turnaround time is under an hour today, and the founders aim to cut it to mere minutes.

Founders Ryan Daniels and John Sarihan say controlling the entire legal workflow, not just the software, is key to real efficiency gains. Sequoia’s partners call contract negotiation a persistent growth bottleneck and see the $300 billion legal market as ripe for direct disruption.

The Atlantic argues that as generative AI floods businesses with infinite content and options, human taste is becoming the defining leadership differentiator. Author Nitin Nohria says discerning which AI-generated variation is most meaningful now rivals vision and execution in importance.

Large language models can churn out essays, logos, and strategies, but the essay stresses they have “no taste” because they cannot originate style or grasp cultural nuance. Nohria highlights Steve Jobs’ meticulously staged iPhone reveal and OpenAI’s purchase of Jony Ive’s startup Io as proof that cultivated taste can elevate technology.

With every product, plan, and pitch now spawning dozens of AI-generated variants, the individual who can choose wisely gains outsized organizational value. The piece warns that absent such judgment, AI systems already show they can generate offensive or biased content, citing a Google shopping tool that produced erotic images of celebrities and minors.