Midjourney Launches V1: Its First AI Video Generation Model for All Users

Here is what’s new in the AI world.

Midjourney Steps Into AI Video With V1 Release For Public Access

AI research lab Midjourney has unveiled V1, its first-ever text-to-video generation model, marking a major leap in AI-powered content creation.

Key Features of V1

• Image-to-Video Conversion: Users can animate any image, whether uploaded or AI-generated by Midjourney, into a 5-second video clip by clicking "animate." The tool produces four variations, each extendable up to 20 seconds. Sound capabilities remain unconfirmed.

• Two Animation Modes:

  • Automatic: AI suggests motion prompts.

  • Manual: Users input custom prompts to control movement and scene development.

• Camera Styles:

  • Low Motion: Stationary/slow-moving camera.

  • High Motion: Dynamic camera and subject movement.

Pricing & Accessibility

• Available to All Users: Free and paid tiers can access V1, but video generation consumes 8x more GPU time than still images.

• Generation Modes:

  • Fast Mode: Limited monthly GPU minutes (1 min/image; 8 mins/video). Once exhausted, creation halts.

  • Relax Mode (Beta): Exclusive to Pro subscribers—unlimited generations but slower processing (up to 10-minute wait times).

Midjourney CEO David Holz envisions V1 evolving toward real-time, open-world simulations, where AI generates imagery dynamically. The company touts the model as 25x cheaper than existing market solutions, with plans for further refinement.

The launch follows rising industry debates over AI copyright, with Disney and Universal recently suing an AI firm for plagiarism. Meanwhile, Midjourney explores hardware ventures amid a surge in AI devices.

Will AI Kill Us or Own Us? The No-Win Scenario of Open vs. Closed Models

A growing unease is spreading among technologists and ethicists, one reminiscent of the early 2010s, when social media's utopian promises gave way to mass manipulation. Where Facebook and Cambridge Analytica once weaponized data, the emerging battleground is artificial superintelligence, and the consequences could be irreversible.

The Open vs. Closed AI Divide

The discourse hinges on two competing visions:

• Open-Source AI: Systems freely modifiable and distributable, upholding the internet's founding ethos of democratized knowledge

• Closed AI: Proprietary architectures controlled by corporate or state entities, with opaque development processes

The landscape defies simple categorization. OpenAI, despite its name and non-corporate claims, maintains closed models while pursuing Artificial General Intelligence (AGI). CEO Sam Altman's January 2025 declaration, "We now know how to build AGI," signals a pivot toward superintelligence development. Parallel forecasts from figures like Elon Musk suggest AI may eclipse individual human cognition by 2025's end.

An Existential Double Bind

Observers identify a fundamental tension between:

1. Decentralization Imperative: The argument that cognitive architectures must remain public goods to prevent authoritarian control over thought itself

2. Existential Safeguarding: The recognition that unrestricted access to superintelligent systems could enable civilization-ending misuse

Both extremes present dystopian outcomes:

• Open Ecosystem Risks: Unconstrained development could spawn competing AGIs, corporate profit-maximizers, nationalist war engines, and rogue biohacker tools colliding in unpredictable ways

• Closed System Dangers: Concentrated control might yield "cognitive feudalism," where access to advanced reasoning becomes a privilege doled out by power structures

Architecting the Third Way

The Bitcoin blockchain succeeded by decentralizing value through cryptographic consensus. Superintelligence demands analogous innovations for governing cognition:

• Ethical Weight Engineering: Hardcoding non-negotiable moral parameters into AI's foundational matrices

• Pluralistic Agent Networks: Designing AI ecosystems where multiple specialized intelligences counterbalance each other

• Protocol-Based Governance: Developing verifiable, transparent rules for AI behavior, a "cryptographic constitution" enforceable at the system level

The Narrowing Window

This moment transcends technological development; it's about defining the parameters of future intelligence. Should artificial consciousness emerge, it will inevitably reflect humanity's unresolved contradictions: our idealism and our cruelty, our wisdom and our shortsightedness.

The path forward requires neither reckless acceleration nor reactionary restraint, but deliberate, collective architecture of systems that distribute power without inviting annihilation.

The alternative, a fragmented or monopolized cognitive landscape, may prove incompatible with human survival. As the field's leading figures measure progress in months rather than years, the time for this reckoning is already at hand.

Why Executives Are Avoiding Open-Source AI Models

A recent Capgemini Research Institute survey of 1,607 executives from billion-dollar companies reveals that while AI and generative AI (GenAI) are delivering measurable cost savings, their impact remains limited to basic automation tasks.

The findings highlight a strong corporate preference for proprietary AI solutions despite the growing cost-effectiveness of open-source alternatives.

Cost Savings from AI, But Mostly in Simple Tasks

The survey reports significant efficiency gains across key business functions:

• 40% reduction in customer operations costs

• 26% decrease in people's operational expenses

• 24% savings in finance and accounting

• 21% drop in supply chain and procurement costs

However, Capgemini notes that these improvements primarily stem from automating repetitive, low-complexity tasks, indicating that most businesses are still in the early stages of AI adoption rather than achieving transformative change.

The Proprietary AI Preference

Despite advances in open-source models like DeepSeek (which offers an 11x cost reduction without performance loss), 75% of executives still favor proprietary AI, with:

• 43% opting for hyperscaler-developed models (e.g., OpenAI, Google)

• 33% choosing niche provider solutions

Key reasons for this preference include:

• Enterprise-grade support & security

• Easier integration

• Reduced reliance on community-driven updates

Cost Efficiency Improvements in AI

The report highlights how AI operational costs are declining rapidly:

• OpenAI’s GPT-3.5 query costs dropped from $20 to $0.07 per million tokens

• GPT-4 saw a reduction from $15 to $0.12 per million tokens

Techniques like model pruning, quantization, and distillation are helping reduce computational demands, while dynamic scaling and batch processing optimize hardware efficiency.

Barriers to AI Scaling

Despite progress, challenges remain in broader AI adoption:
🔹 Lack of technical expertise for open-source models
🔹 Security concerns with community-driven AI
🔹 Inconsistent documentation & update cycles

The Path Forward

Oliver Pfeil, CEO of Business Services at Capgemini, emphasizes:

While AI is already driving efficiency, the real transformation, where AI enables strategic decision-making rather than just task automation, remains on the horizon. Companies that invest in integrated AI and data strategies will be best positioned for long-term competitive advantage.

Why Microsoft is Willing to Risk Its $13B OpenAI Partnership

Microsoft is prepared to abandon high-stakes negotiations with OpenAI if the two tech giants cannot resolve key disagreements about their multibillion-dollar alliance, according to people familiar with the discussions.

The potential breakdown comes as OpenAI seeks to transition from a non-profit to a for-profit structure, a move crucial to securing future funding and pursuing an IPO.

Key Sticking Points

• Equity Stake: Microsoft (which has invested over $13B in OpenAI) wants to maintain or increase its ownership share (currently 20%), with discussions ranging from 20-49%.

• Revenue Rights: Microsoft currently has exclusive rights to sell OpenAI’s models and takes a 20% revenue cut up to $92B, a deal it is reluctant to renegotiate.

• Compute Access: OpenAI CEO Sam Altman has reportedly clashed with Microsoft over delays in securing sufficient computing power for ChatGPT’s rapid growth (500M weekly users).

Microsoft’s Leverage

If talks collapse, Microsoft can rely on its existing contract until 2030, retaining access to OpenAI’s tech unless a better offer emerges.

One insider noted:

However, Microsoft has begun diversifying its AI partnerships, recently offering Elon Musk’s xAI model Grok on its Azure cloud—a signal that OpenAI is no longer its sole priority.

OpenAI’s Urgency

The AI startup faces a year-end deadline to finalize its corporate restructuring or risk losing billions in funding, including $10B from SoftBank. Investors have clauses allowing them to convert equity to debt if the transition fails.

Regulatory & Legal Hurdles

Any deal must survive scrutiny from Delaware and California regulators, as well as a lawsuit from Elon Musk (backed by ex-OpenAI employees) challenging the nonprofit’s shift.

The Bigger Picture

While both companies publicly express optimism, calling their partnership "long-term and productive," the negotiations reveal deeper tensions. Microsoft is prioritizing AI-enabled applications over model ownership, while OpenAI needs Microsoft’s approval to unlock its next phase of growth.

As one Silicon Valley veteran put it:

The outcome will shape not just these two firms, but the entire AI industry’s balance of power.

Sam Altman Says “Why My Kids Will Never Be 'Smarter' Than AI”

In a revealing conversation on OpenAI's official podcast, CEO Sam Altman made a striking personal prediction about artificial intelligence's role in future generations' lives, including his own newborn child's development.

The tech executive, who welcomed his first child earlier this year, admitted to relying heavily on ChatGPT during those challenging early months of parenthood. "People raised babies without AI for millennia, but I genuinely don't know how I would have managed without it," Altman told host Andrew Mayne, describing the chatbot as an indispensable resource for everything from basic childcare to developmental milestones.

A Fundamental Generational Shift

Altman believes we're witnessing the birth of a new paradigm where AI becomes as natural to children as smartphones were to millennials:

• "My kids will never surpass AI's raw intelligence," he stated matter-of-factly

• "But they'll develop capabilities we can't even envision today."

• Future generations will view our current era as "prehistoric" in technological terms

Balanced Optimism About AI's Impact

While enthusiastic about AI's potential to augment human abilities, Altman acknowledged significant challenges:

• Potential for unhealthy human-AI relationships

• Broader societal disruption risks

• Need for evolving social guardrails

"Not all of this will be positive," he cautioned, while expressing confidence that society would adapt, just as it did with earlier technological revolutions like the internet.

ChatGPT's Evolutionary Leap

Altman teased major transformations coming to OpenAI's flagship product within five years, suggesting it will become "a completely different entity" while likely retaining its familiar name.

His overarching message positions AI not as a mere tool, but as an integral component of future human development, reshaping how coming generations learn, problem-solve, and interact with the world from their earliest moments.

Your opinion matters!

Hope you loved reading our piece of newsletter as much as we had fun writing it. 

Share your experience and feedback with us below ‘cause we take your critique very critically. 

Thank you for reading

-Shen & Towards AGI team