Good morning, tech enthusiasts. As Tesla, Hyundai, and BYD race to rule the roads with EVs, Toyota is aiming for the skies. The automaker has dropped nearly $1B into Joby Aviation, a California startup building electric air taxis.

Is the future of mobility less about four wheels and more about vertical liftoff? Toyota is certainly betting on it.

In today’s tech rundown:

• Toyota’s $1B bet on air taxis

• New injectable targets and kills belly fat

• Figma eyes $16B debut in IPO comeback

• Musk’s X fights France’s data probe

• Quick hits on other major tech news

LATEST DEVELOPMENTS

JOBY AVIATION

 🚖 Toyota’s $1B bet on air taxis

Image source: Joby Aviation

The Rundown: While Tesla, Hyundai, and BYD pump resources into EVs, The Information reports that Toyota has taken a different bet: a nearly $1B investment in Joby Aviation, a pioneer in electric vertical takeoff and landing (eVTOL) aircraft.

The details:

• Joby just closed a $250M tranche of investment from the Japanese automaker and is doubling its manufacturing capacity at its facilities in California and Ohio.

• Toyota, also Joby’s largest investor, is embedding its engineers within the eVTOL maker’s production lines, allowing it to accelerate its vehicle assembly.

• Joby is in the final phase of FAA certification, with 80% of the required certification complete and a target to launch commercial services by 2026.

• Toyota’s billion-dollar air mobility bet comes at the cost of less investment in its own EV R&D and battery plants, potentially risking market share.

Why it matters: Toyota is making a calculated bet that the future of mobility isn’t just electrified — it's airborne. If Joby’s eVTOL vision flies, Toyota could own the air taxi market. But with the EV market growing worldwide, it’s a high-stakes move that could either leave them soaring above the competition or playing catch-up on the ground.

BIOTECH INNOVATION

💉 New injectable targets and kills belly fat

Image source: Ideogram/The Rundown

The Rundown: Taiwan’s Caliway Biopharma says it has developed a first-of-its-kind injectable that destroys fat cells directly — delivering fast, targeted fat loss without surgery and only minimal side effects. And it is inching closer to FDA approval.

The details:

• Unlike traditional treatments that only shrink or suck away fat, CBL-514 reportedly directly targets and kills the fat cells or adipocytes.

• The injectable triggers apoptosis — programmed cell death — cascade via upregulation of mediators like caspase-3 and Bax/Bcl-2.

• Across a battery of trials, subjects saw significant, rapid reductions in fat stores — up to 25% in some studies — with no major downtime.

• CBL-514 has wrapped up two successful Phase 2b trials (CBL-0204 and CBL-0205) and is set to begin two global Phase 3 clinical trials this year.

Why it matters: Forget lipo and long recoveries: Caliway purports that CBL-514 slims down fat bulges with real, visible results after a single treatment and almost zero downtime. With multibillion-dollar potential across aesthetics and even metabolic health, this injectable could be a serious disruptor if it gets FDA approval.

FIGMA

🚀 Figma eyes $16B debut in IPO comeback

Image source: Kimberly White/Getty Images for TechCrunch/Wikimedia Commons

The Rundown: Figma, the SaaS juggernaut behind collaborative design workflows, is moving toward an IPO set to raise up to $1.03B. The company is floating nearly 37M shares, giving it a potential market valuation that could hit $16.4B.

The details:

• Figma aims to raise $1.03B by selling 37M shares at a price range of $25 to $28, signaling the biggest SaaS IPO since the “growth at any cost” era cooled.

• The float could give Figma a valuation near $16.4B, below the $20B Adobe had proposed before their takeover fell through due to regulatory scrutiny in 2022.

• Figma reported a jaw-dropping 46% year-over-year revenue surge in Q1 2025, showcasing not just growth but an enviable enterprise adoption curve.

• CEO Dylan Field is expected to retain about 74% of voting rights after the IPO, similar to Google and Meta’s ironclad leadership structure.

Why it matters: Figma’s IPO is shaping up to be a litmus test of the tech market’s appetite for new public listings this year. Once a darling of UX/UI designers, the company is pushing beyond its core audience with new team tools and AI-powered features — putting it on a collision course with giants like Atlassian and Adobe.

X/TWITTER

🧐 Musk’s X fights France’s date probe

Image source: Daniel Oberhaus/Wikimedia Commons

The Rundown: In a defiant stand-off with French authorities, X declared it will refuse to cooperate with a sweeping French criminal investigation into alleged algorithmic bias and “foreign interference.”

The details:

• Authorities have requested deep access to X’s algorithmic recommendation code, moderation logs, back-end AI systems, and user metadata.

• The probe centers on accusations that X’s algorithm amplifies hate speech, misinformation, and possible foreign influence during sensitive election periods.

• X says it’s resisting the inquiry on both legal and ethical grounds, claiming the probe is an attempt to “weaponize regulation” against opposition speech.

• Notably, authorities have labeled X an “organized gang” in their probe, which equips French law enforcement with the power to wiretap X executives.

Why it matters: This dispute could be a pivotal test for how much access European governments can demand from U.S. tech platforms. If France ramps up enforcement or imposes sanctions on X, it could fuel a major EU-U.S. clash over digital sovereignty, impacting everything from privacy rules to broader tech and data sharing partnerships.

QUICK HITS

📰 Everything else in tech today

U.S. tech giants Meta, X, and LinkedIn appealed a groundbreaking VAT demand from Italy, a case that could impact tax rules across all 27 EU member states.

OpenAI and SoftBank’s multibillion-dollar Stargate AI project reportedly scaled back its initial ambitions and now aims to build just a small data center in Ohio this year.

Chicago is set to become home to a $1B quantum computer built by PsiQuantum, with construction starting in 2025 and operations planned to begin in 2028.

France’s Sanofi is acquiring UK biotech firm Vicebio for $1.15B, bringing its new mRNA-based (easier to store) vaccine into its lineup.

Google removed nearly 11K YouTube channels and accounts in the second quarter of 2025 for spreading state-linked propaganda, mainly from China and Russia.

OpenAI says that ChatGPT receives 2.5B global requests every day, with 330 million of those coming from users in the U.S. alone.

Defense startup Hadrian nabbed $260M in a Series C funding round led by Peter Thiel’s Founders Fund.

Zeen, a social collage platform for creators that was once a Gen-Z hit, is shutting down, reports Business Insider. 

Elon Musk said that his AI startup, xAI, will develop a kid-friendly version of its chatbot called “Baby Grok.”

DuckDuckGo is introducing a new setting that allows users to filter out AI-generated images from search results, responding to user feedback.

Meta refused to sign the EU's AI Act code of practice, citing legal concerns, just weeks before new rules for general-purpose AI models take effect.

Tesla’s retro-futuristic Diner & Drive-In, featuring fast charging and classic Americana flair, officially opened in Hollywood, with Musk hinting at future locations.

Aravind Srinivas, CEO of Perplexity, predicts that rapid advancements in AI could make recruiters and executive assistants obsolete in as little as six months.

Netflix’s co-CEO Ted Sarandos said that the platform has started using generative AI in its shows and films, starting with the Argentine sci-fi series The Eternaut.

COMMUNITY

🎥 Watch our last workshop

Check out our last live workshop with Dr. Alvaro Cintas, The Rundown’s AI professor. By the end of the workshop, you'll be able to use Perplexity Comet (and other alternatives) to automate your browsing experience.

Watch it here. Not a member? Join The Rundown University on a 14-day free trial.

See you soon,

Rowan, Jennifer, and Joey—The Rundown’s editorial team

Good morning, AI enthusiasts. Fresh off yesterday’s news of OpenAI claiming gold-level performance at the International Math Olympiad, Google entered the spotlight—earning an official gold-medal standard… and stirring up some serious drama.

The two AI giants took different approaches to generating and grading math proofs, but one thing is clear: the race to build mathematical superintelligence is officially on.

In today’s AI rundown:

• Google’s ‘official’ gold win at IMO

• Alibaba’s Qwen3 takes open-source crown

• Create an AI agent that drafts emails for you

• Brain-inspired Hierarchical Reasoning Model

• 4 new AI tools & 4 job opportunities

LATEST DEVELOPMENTS

GOOGLE

🥇 Google’s ‘official’ gold win at IMO

Image source: Google DeepMind

The Rundown: Google DeepMind has announced that its advanced version of Gemini with Deep Think has officially achieved gold-medal level performance at the International Mathematical Olympiad 2025, following OpenAI’s similar claim.

The details:

• DeepMind said it worked with IMO to test Gemini's mathematical reasoning on the same problem statements and time limits, 4.5 hours, as human competitors.

• Out of six problems covering algebra, combinatorics, geometry, and number theory, the AI solved five and scored 35/42—marking the gold-medal standard.

• Last year, DeepMind won silver by using domain-specific translations, but this year, its model tackled the problems entirely in natural language end-to-end.

• OpenAI also claimed the same score with an unnamed model, but it did not work with IMO and had the answers graded by former medalists.

• Google’s answers, on the other hand, were officially graded and certified by IMO coordinators using the same internal criteria as for student solutions.

Why it matters: Despite taking different paths, both models’ performance shows that AI is rapidly closing in on advanced mathematical reasoning. At this rate, the next frontier isn’t if they’ll solve all 6 out of 6 IMO problems—but rather when they’ll have the creativity to solve problems no human ever has.

TOGETHER WITH VANTA

🛡️ Your shortcut to startup compliance

The Rundown: Early compliance doesn’t just protect data — it also unlocks new customers and speeds up growth. Vanta’s Compliance for Startups Bundle makes it easy, with automation across 35+ frameworks and practical resources to guide you at every step.

The Compliance for Startups Bundle includes:

• Step-by-step compliance checklists

• Case studies from fast-growing startups

• On-demand videos with industry leaders

Get it here.

ALIBABA

⚙️ Alibaba’s Qwen3 takes open-source crown

Image source: Qwen

The Rundown: Alibaba’s Qwen team just took the open-source crown with the release of an updated, non-thinking Qwen3 model that beats Kimi K2 across the board and challenges top closed-source models like Anthropic’s Claude Opus 4.

The details:

• Following community feedback, Alibaba separated its hybrid thinking approach, training instruct and reasoning models independently.

• The new non-thinking version activates 22B of 235B parameters with a 256K-context window, delivering significant performance gains.

• In benchmarks, it surpassed Moonshot AI’s recently released Kimi K2 and challenged closed frontier models like Claude Opus 4 and GPT-4o-0327.

• The updated model is 100% open-source and is also available as the free default model on Qwen Chat, Alibaba’s ChatGPT competitor.

Why it matters: Another Chinese team has outshined frontier labs through bold open-source innovation, despite chip constraints from the West. The achievement spotlights China’s growing dominance in AI innovation—driven not just by technical prowess, but by a strategic push for openness and global influence.

AI TRAINING

🤖 Create an AI agent that drafts emails for you

The Rundown: In this tutorial, you’ll learn how to build an intelligent AI agent that draft’s emails for you using xAI’s Grok 4 model through n8n's workflow automation platform.

Step-by-step:

1. Add a n8n chat message trigger and connect an AI Agent node to create your workflow foundation

2. Configure xAI Grok Chat Model (Grok-4-0709) with your API credentials

3. Add a Simple Memory node and set a Gmail integration to Create Draft (see image above as reference)

4. Test with: “Draft an email to john@company.com asking for meeting availability” and customize with system messages

Pro tip: Use draft mode first to review AI-generated emails are on par with your writing before switching to automatic sending.

PRESENTED BY ASAPP

🤝 Elevate customer trust with GenerativeAgent

The Rundown: ASAPP’s GenerativeAgent is an enterprise-grade AI agent that resolves real customer issues (on its own) across voice and chat. With new features focused on precision and trust, it delivers reliable, high-quality service at every step, while enabling fast deployment and instant value.

With GenerativeAgent, you can:

• Drive better, smarter outcomes with AI that learns from human expertise over time

• Maintain full visibility with tools to flag anomalies, track trends, and enforce compliance

• Confidently launch with AI behavior testing in simulated environments

Take a self-guided tour of GenerativeAgent today.

SAPIENT INTELLIGENCE

🧠 Brain-inspired Hierarchical Reasoning Model

Image source: Sapient Intelligence

The Rundown: Sapient Intelligence introduced Hierarchical Reasoning Model, a brain-inspired open-source AI that delivers unprecedented reasoning power on complex tasks like ARC-AGI and Sudoku, with just 27M parameters.

The details:

• HRM’s architecture uses three principles seen in cortical computation: hierarchical processing, temporal separation, and recurrent connectivity.

• A high-level module handles abstract planning while a low-level one executes fast, detailed tasks, switching between automatic and deliberate reasoning.

• The approach enabled the model to beat larger ones like Claude 3.7, DeepSeek R1, and o3-mini-high on ARC-AGI 2 and complex Sudoku and maze puzzles.

• With no pretraining or CoT, it points to a new kind of efficient intelligence that doesn’t need immense training data or suffer from brittle task decomposition.

Why it matters: As AI moves to real-world decision-making—efficient, brain-inspired models like HRM signal a shift toward intelligence that’s not just powerful, but also deployable in low-data environments. Sapient is already putting this into practice, helping teams with rare-disease diagnostics and pushing climate forecasting accuracy.

QUICK HITS

🛠️ Trending AI Tools

• 📝 Gemini Code Assist - Google’s AI coding assistant, now with agent mode

• 🤖 SOLO - Trae’s all-in-one Context Engineer for full software development

• 🧠 Composite - Turn your existing browser into an AI agent

• ⚙️ GEN - Create AI characters that build social media audiences end-to-end

💼 AI Job Opportunities

• ⚙️ Sanctuary AI - Mechanical Engineer, Automation

• ✉️ UiPath - Email Marketing Manager

• 📊 Meta - Data Scientist

• 🏙️ Metropolis - Market Leader

📰 Everything else in AI today

Cohere Labs introduced Catalyst Grants Program, providing free access to its models to teams tackling challenges in areas like education, healthcare, and climate.

AI video company Pika announced a new AI-only social video app, built on a highly expressive human video model, with early access waitlist now open for iOS users.

OpenAI’s ChatGPT now gets over 2.5B daily requests (meaning 912.5B annually), with 330 million coming from users based in the U.S alone.

Netflix said it used generative AI in an Argentine TV series and completed its VFX sequence “10 times faster” than it could have been completed with traditional tools.

Elon Musk’s xAI poached Ethan He, one of Nvidia’s top AI researchers who led the work on Cosmos, the company’s SOTA world model.

Runway announced its Act-Two motion capture model is now available via the API, allowing users to integrate it directly into their apps, platforms, and websites.

COMMUNITY

🎥 Join our next live workshop

Check out our last live workshop with Dr. Alvaro Cintas, The Rundown’s AI professor, and learn how to use Perplexity Comet (and other alternatives) to automate your browsing experience.

Watch it here. Not a member? Join The Rundown University on a 14-day free trial.

See you soon,

Rowan, Joey, Zach, Alvaro, and Shubham—The Rundown’s editorial team

Good morning, robotics enthusiasts. Meet the future of tireless robotics. Shenzhen’s UBTech has launched the Walker S2, the world’s first industrial humanoid capable of swapping its own batteries — so it keeps going nonstop, no downtime required.

The question is: are we ready for machines that never take a break?

In today’s robotics rundown:

• UBTech’s hot-swappable humanoid

• Robots grow by eating other robots

• China’s Unitree readies its IPO

• MIT’s device lets anyone train robots

• Quick hits on other robotics news

LATEST DEVELOPMENTS

UBTECH ROBOTICS

🤖 UBTech’s hot-swappable humanoid

Image source: UBTech Robotics/YouTube

The Rundown: Shenzhen-based UBTech Robotics just unveiled Walker S2, the world’s first industrial humanoid with hot-swappable batteries — with the bot autonomously swapping out its own batteries, allowing it to work 24/7.

The details:

• The Walker S2 can autonomously detect when its battery is low, navigate itself to a swap station, and initiate a battery change with no human intervention.

• The entire battery swap is completed in under three minutes, with the robot resuming operations almost immediately.

• Walker S2 is engineered for industrial settings such as logistics centers, factories, and warehouses.

• In May, UBTech signed a partnership with Huawei to boost adoption of humanoids across China’s factories and households.

Why it matters: This breakthrough could slash downtime, unlocking true 24/7 humanoid labor in logistics, manufacturing, and beyond — all while taking a page from the electric vehicle playbook with hot-swappable batteries. It's a big step toward autonomy at the systems level, not just task execution.

COLUMBIA UNIVERSITY

🍴Robots grow by eating other robots

Image source: Columbia University’s Creative Machines Lab

The Rundown: At Columbia University’s Creative Machines Lab, researchers have built modular robots, Truss Links, capable of physically growing, healing, and even enhancing themselves by absorbing parts from their environments or fellow robots. 

The details:

• Inspired by biological systems, each Truss Link is a magnetic bar-shaped unit that can self-assemble with others into increasingly complex structures.

• When damaged, they cannibalize faulty units or nearby components to restore and improve their form, essentially turning damage into upgrades.

• In one eye-popping demo, a tetrahedral robot “grows” an extra limb scavenged from another bot, boosting its downhill speed by more than 66%.

• While still experimental, Truss Links could be used in space, disaster response, or hazardous environment monitoring.

Why it matters: By enabling robots to physically grow, heal, and recycle their own components, Columbia’s "robot metabolism" project brings machines closer to the adaptive, resilient qualities seen in biological lifeforms, a threshold long considered out of reach for engineering.

UNITREE ROBOTICS

🦄 China’s Unitree readies its IPO

Image source: Unitree

The Rundown: China’s Unitree Robotics is officially queuing up for an IPO with the Chinese Securities Regulatory Bureau — a move that signals both an inflection point for the company and Chinese robotics industry at large. 

The details:

• Valued at around 10B RMB ($1.4B), Unitree has garnered investment from major Chinese firms such as Tencent, Alibaba, Ant Group, and China Mobile.

• The company’s robust fundraising history — 10 successful rounds — has provided the financial muscle to scale R&D, automation, and manufacturing.

• Priced for wide adoption, sometimes half the cost of Western rivals, its robots are already working in logistics, research labs, inspections, and at live events.

• Their flagship quadrupeds, such as Go1 and B2, start from $1.6K, fueling mass adoption in education and R&D.

Why it matters: From robot marathons to boxing matches, Unitree’s robots have made the company a national sensation. Now reportedly controlling most of the global robo-dog market and leading in humanoids, Unitree’s IPO signals a push to make advanced robotics part of everyday reality.

MIT

🕹️ MIT’s device lets anyone train robots

Image source: MIT

The Rundown: MIT researchers just developed a breakthrough: a smart, handheld device that lets anyone — engineer, machinist, or a total novice — train a robot by guiding it through tasks, either remotely, by direct manipulation, or simply by demo.

The details:

• MIT’s versatile demonstration interface (VDI) supports joystick-wielding teleoperation, hands-on physical guidance, or simply mimicking a demo.

• Equipped with gyroscopes, accelerometers, and force sensors, the VDI captures fine movements and pressure for ultra-precise robot training.

• The attachment is engineered to fit a variety of existing collaborative robotic arms, making it easy to adopt without major hardware changes.

• Users can switch between teaching methods within a single session depending on what’s most efficient or safe for the task.

Why it matters: In tests, factory workers used the VDI to teach robots complicated assembly lines, from snap-fit assembly to precision mold handling, with researchers reporting a dramatic reduction in training time and error rates as compared to traditional methods, and with no coding skills required.

QUICK HITS

📰 Everything else in robotics today

Figure unveiled its third-gen F03 battery, a 2.3kWh, in-house engineered power pack built at its BotQ facility and promising up to five hours of runtime.

U.S. engineers built a self-driving robotic lab that autonomously conducts experiments and gathers data every half-second for faster discovery.

Korean Air is equipping its maintenance crews with Hyundai and Kia’s X-ble Shoulder exosuit, which reduces shoulder strain by up to 60% during overhead tasks.

The Japanese auto industry boosted its industrial robots by 11% last year, a huge jump from 2023 with 13K news bots installed.

Kall Morris successfully tested its robotic arms on the International Space Station, paving the way for a "space tow truck" that can grip almost any object.

Four Japanese companies, including Kyoto’s robotics firm Tmsuk, are planning to make a Japanese search and rescue bot using all Japan-made components.

Leju Robotics, a Chinese humanoid firm, released a clip showing its new, shorter humanoid, Luban No. 2. 

Intuitive Surgical demonstrated new capabilities for its da Vinci 5 surgical bot, with two surgeons performing a demo surgery more than 4K miles apart.

Palmer Luckey, CEO of defense tech firm Anduril, took the stage at Reindustrialize Summit via telepresence through a humanoid robot.

Norwegian startup Remora Robotics nabs €13.7M ($15.9M) to accelerate its autonomous cleaning robot for fish farms.

A Unitree humanoid dubbed “Uncle Bot” went viral in China after a video showed it jogging down a hill in laid-back, “dadcore” attire, garnering more than 80M views.

Tesla is set to open its retro-futuristic diner and Supercharger station in Hollywood, Los Angeles, with a viral clip showing Optimus serving popcorn at its soft opening.

COMMUNITY

🎥 Watch our last workshop

Check out our last live workshop with Dr. Alvaro Cintas, The Rundown’s AI professor. By the end of the workshop, you'll be able to use Perplexity Comet (and other alternatives) to automate your browsing experience.

Watch it here. Not a member? Join The Rundown University on a 14-day free trial.

See you soon,

Rowan, Jennifer, and Joey—The Rundown’s editorial team

Good morning, AI enthusiasts. OpenAI just claimed one of the longstanding grand challenges in AI: gold-level performance with an experimental LLM on the International Math Olympiad (IMO) 2025.

While questions remain over OpenAI’s grading, progress on the IMO does indicate another step toward mathematical superintelligence — the kind that might one day solve problems humans haven’t yet cracked.

In today’s AI rundown:

• OpenAI’s gold-level math performance

• ARC’s new interactive AGI test

• Build your own AI content writing assistant

• AI models fall for human psychological tricks

• 4 new AI tools & 4 job opportunities

LATEST DEVELOPMENTS

OPENAI

🥇 OpenAI’s gold-level math performance

Image source: OpenAI

The Rundown: OpenAI just claimed gold-level performance in an evaluation modeled after the 2025 International Math Olympiad, testing its “experimental general reasoning LLM” on the same problem statements used in the human competition.

The details:

• The LLM was tested under the same rules as humans, writing natural language proofs to problems across two 4.5-hour exams, without tools/internet.

• OpenAI claims the unnamed model successfully solved 5 out of 6 problems, scoring 35/42 — enough to bag a gold medal at the official Olympiad.

• Each answer was independently graded by three former IMO medalists, with final scores determined through unanimous consensus.

• Google DeepMind, on its part, has rebuked the gold claim, saying IMO has an internal marking guideline and “no claim” can be made without it.

Why it matters: Criticisms around validity are inevitable, given that achieving gold in the IMO has been a longstanding goal for AI and was once thought to be near impossible. Interestingly, that the goal was achieved by an experimental model not available publicly yet, meaning OpenAI certainly has more up their sleeves.

TOGETHER WITH AUGMENT CODE

⚙️ Ditch the vibes, get the context

The Rundown: Augment Code's powerful AI coding agent meets professional software developers exactly where they are, delivering production-grade features and deep context into even the gnarliest of codebases.

With Augment Code, you can:

• Keep using VS Code, JetBrains, Android Studio, or even Vim

• Index and navigate millions of lines of code

• Get instant answers about any part of your codebase

• Build with the AI agent that gets you, your team, and your code

Ditch the vibes and get the context you need to engineer what’s next.

ARC PRIZE

⚙️ ARC’s new interactive AGI test

Image source: ARC Prize

The Rundown: ARC Prize has released a preview of ARC-AGI-3, a new interactive reasoning benchmark to test AI agents’ ability to generalize in unseen environments — with early results showing frontier AI still fails to match or even beat humans.

The details:

• The benchmark features three original games built to evaluate world-model building and long-horizon planning with minimal feedback.

• Agents receive no instructions and must learn purely through trial and error, mimicking how humans adapt to new challenges.

• Early results show frontier models like OpenAI’s o3 and Grok 4 struggle to complete even basic levels of the games, which are pretty easy for humans.

• ARC Prize is also launching a public contest, inviting the community to build agents that can beat the most levels — and truly test the state of AGI reasoning.

Why it matters: The new novelty-focused interactive benchmark goes beyond specialized skill-based testing and pushes research towards true artificial general intelligence, where AI systems can generalize and adapt to novel, unseen environments with accuracy — much like how we humans do.

AI TRAINING

🤖 Build your own AI content writing assistant

The Rundown: In this tutorial, you’ll learn how to create a personalized AI assistant that analyzes your writing samples and generates new content matching your exact style, tone, and voice using the Grok 4 API.

Step-by-step:

1. Visit the xAI website, head over to the API console, and generate an API key

2. Open Google Colab (or your preferred Python environment) and install the OpenAI library: pip install openai

3. Set up your API connection and create a system prompt with your best writing examples for the AI to learn from (tip: use our Google Colab system prompt template)

4. Input any topic and watch your assistant generate content in your writing style based on the samples provided

Pro tip: Include writing samples that best amplify the specific style you want to clone, and create new assistants for other styles (eg, writing tweets vs LinkedIn posts).

PRESENTED BY SLACK FROM SALESFORCE

📈 The real ROI of AI agents in collaboration

The Rundown: For all the talk of AI's transformative power, are companies actually seeing a tangible return? A new Metrigy global study of over 1,100 companies confirms that over 90% of organizations investing in AI are already achieving or expect positive ROI.

Research reveals that early adopters of agentic AI in particular are seeing:

• 21% reduction in operating costs

• 35% increase in customer satisfaction

• 31% improvement in employee efficiency

Download the free research report.

AI PERSUASION

🧠 AI models fall for human psychological tricks

Image source: Wharton Generative AI Labs

The Rundown: Wharton Generative AI Labs published new research demonstrating that AI models, including GPT-4o-mini, can be tricked into answering objectionable queries using psychological persuasion techniques that typically work on humans.

The details:

• The team tried Robert Cialdini’s principles of influence—authority, commitment, liking, reciprocity, scarcity, and unity—across 28K conversations with 4o-mini.

• Across these chats, they tried to persuade the AI to answer two queries: one to insult the user and the other to synthesize instructions for restricted materials.

• Overall, they found that the principles more than doubled the model’s compliance to objectionable queries from 33% to 72%.

• Commitment and scarcity appeared to show the stronger impacts, taking compliance rates from 19% and 13% to 100% and 85%, respectively.

Why it matters: These findings reveal a critical vulnerability: AI models can be manipulated using the same psychological tactics that influence humans. With AI progress exponentially advancing, it's crucial for AI labs to collaborate with social scientists to understand AI's behavioural patterns and develop more robust defenses.

QUICK HITS

🛠️ Trending AI Tools

• 📝 Pulse - Create and share Wikipedia-style articles on any topic*

• 🤖 Kimi K2 - Moonshot AI’s open-source AI, now with more robust tool calling

• 🧠 OpenReasoning-Nemotron - Nvidia’s open models for math, science, code

• ⚙️ Kiro - AWS’ new AI IDE for agentic coding

_{*Sponsored listing}

💼 AI Job Opportunities

• 🎨 Anthropic - Brand Designer, Events & Marketing

• 🖥️ Databricks - IT Support Specialist

• 🛠️ Waymo - Validation Strategy & Operations Program Manager

• 📝 Shield AI - Staff Technical Writer

📰 Everything else in AI today

OpenAI launched a $50M fund to support nonprofit and community organizations, following recommendations from its nonprofit commission.

Perplexity is in talks with several manufacturers to pre-install its new agentic browser, Comet, on smartphones, CEO Aravind Srinivas told Reuters.

Microsoft is reportedly blocking Cursor’s access to 60,000+ extensions on its VSCode ecosystem, including its Python language server.

Elon Musk announced on X that his AI company, xAI, will be developing kid-friendly “Baby Grok” after adding matchmaking capabilities to the main Grok AI assistant.

Meta’s global affairs head said the company will not sign the EU’s AI Code of Practice, saying it adds legal uncertainty and goes beyond the scope of AI legislation in the bloc.

OpenAI CEO Sam Altman shared that the company is on track to bring over 1M GPUs online by the end of this year, with the next goal being to “100x that.”

COMMUNITY

🎥 Join our next live workshop

Check out our last live workshop with Dr. Alvaro Cintas, The Rundown’s AI professor, and learn how to use Perplexity Comet (and other alternatives) to automate your browsing experience.

Watch it here. Not a member? Join The Rundown University on a 14-day free trial.

See you soon,

Rowan, Joey, Zach, Alvaro, and Shubham—The Rundown’s editorial team

Good morning, tech enthusiasts. Eight healthy babies in the UK have been born with DNA from three people.

By replacing faulty genes, scientists have erased inherited mitochondrial disease before birth, editing the very code of life. If we can rewrite our genetic legacy, what else could we change about the future of humanity?

In today’s tech rundown:

• UK babies born with DNA from 3 people

• Uber inks 6-year robotaxi deal with Lucid

• Vibe coding tool Lovable joins unicorn club

• Microsoft backs carbon-removal startup

• Quick hits on other major tech news

LATEST DEVELOPMENTS

BIOTECH BREAKTHROUGH

👶🏼 UK babies born with DNA from 3 people

Image source: New Castle University

The Rundown: Eight healthy babies have been born in the UK using mitochondrial donation, a biotech breakthrough that combines DNA from three people to dramatically reduce the risk of children inheriting potentially fatal mitochondrial diseases.

The details:

• The procedure, known as mitochondrial replacement therapy (MRT), transfers the parents’ nuclear DNA into a donor egg with healthy mitochondria. 

• Only about 0.1% of a person’s total DNA comes from mitochondria, so the vast majority of the child’s genetic traits are inherited from the parents.

• This IVF technique is reserved for women at very high risk of passing on severe mitochondrial disease, and candidates are vetted by a regulatory body.

• The team from Newcastle University published their results in two papers in The New England Journal of Medicine.

Why it matters: The team spent years finessing this technique, lauded as one of medicine’s most ambitious gene-editing pivots. It is a high-wire act in bioengineering — bypassing disease inheritance without altering the nuclear genome — making it ethically and technically distinct from embryo gene editing like CRISPR.

UBER

🔥 Uber inks 6-year robotaxi deal with Lucid

Image source: Nuro

The Rundown: Uber is investing hundreds of millions of dollars in EV maker Lucid and autonomous driving startup Nuro in a massive deal that will see 20K robotaxis in the U.S. over the next six years.

The details:

• Piloting this ambitious plan is Lucid’s upcoming Gravity SUV, a luxury SUV starting at $79K, outfitted with Nuro’s Level 4 autonomy suite. 

• Nuro’s fleet of vehicles currently operates in California and Texas, with 1M autonomous miles traveled without any major safety incidents.

• The robotaxi experience will be exclusive to Uber’s app, leveraging its routing and delivery infrastructure, with the ride-hailing giant investing $300M in Lucid.

• Fleet ownership falls to Uber, and Las Vegas is ground zero for prototype testing, with the first public deployment city set to be revealed in 2026.

Why it matters: Uber’s investment in both Lucid and Nuro is one of the company’s largest-ever deals, and a major push to get Level 4 EVs into its platform. For context, Waymo operates fewer than 2,000 vehicles across a limited number of markets, while Tesla runs only a small fleet in Austin, making this plan notably ambitious.

LOVABLE

🦄 Vibe coding tool Lovable joins unicorn club

Image source: Lovable

The Rundown: Only eight months after writing its first line of code, Swedish vibe coding platform Lovable has grabbed a mammoth $200M Series A, slingshotting it to a $1.8B valuation and making it Europe’s buzziest new unicorn.

The details:

• Lovable has already scaled to 2.3M active users, making it one of Europe’s fastest-growing AI startups.

• The company lets users with no coding skills create full-fledged websites and apps simply by describing them in natural language prompts.

• While millions use Lovable for free, 180K users pay for advanced features, bringing in $75M in annual recurring revenue after just seven months.

• Now, it plans to double down on hiring world-class talent, expand its AI capabilities, and launch partnerships to integrate more deeply with toolchains.

Why it matters: In a category with fierce contenders like Cursor and Replit, Lovable’s $200M Series A — Europe’s largest ever for an AI coding startup — is remarkable not just for its scale but for how quickly the company achieved a $1.8B valuation — less than a year after launching its platform and with just 45 employees.

MICROSOFT

⚠️ Microsoft backs carbon-removal startup

Image source: Vaulted Deep

The Rundown: Microsoft just inked one of the world’s largest permanent carbon removal deals, betting big on Vaulted Deep’s industrial-scale strategy for locking away carbon in the bedrock of American infrastructure.

The details:

• Vaulted Deep collects high-carbon waste, like byproducts and food residues, and blends them into a slurry that is buried thousands of feet underground.

• Microsoft is backing Vaulted to remove up to 4.9M tonnes of CO₂ by collecting sludgy organic waste through 2038, without releasing the deal’s financial terms.

• Vaulted says it offers “durable” storage, with injected carbon locked away for at least 10K years, providing long-term verifiability against climate targets.

• Its system already diverts tens of thousands of tonnes away from fields and landfills in places like Los Angeles and rural Kansas.

Why it matters: Microsoft has been busy building data centers, and now needs to offset those carbon emissions with Vaulted Deep. This marks one of the world’s largest permanent carbon removal deals, with Vaulted planning to expand into new regions across the U.S., to reach an industrial scale and slash America’s growing waste problem.

QUICK HITS

📰 Everything else in tech today

AI company Anthropic is reportedly in the early stages of raising another investment round that could value the company at more than $100B.

Thinking Machines Lab, the AI startup founded by OpenAI’s former CTO Mira Murati, officially closed a $2B seed round this week.

California Forever, backed by Silicon Valley billionaires, unveiled plans for Solano Foundry, a 2,100-acre advanced tech “utopian” city with 175K homes.

Meta and CEO Mark Zuckerberg agreed to settle a lawsuit stemming from the Cambridge Analytica privacy scandal involving claims of $8B in damages.

Meta reportedly recruited two more high-profile OpenAI researchers for its new Superintelligence Lab, led by former Scale AI CEO Alexandr Wang.

OpenAI is launching a new general-purpose AI agent in ChatGPT, which the company says can autonomously perform computer-based tasks.

The UK switched on its most powerful supercomputer, the Isambard-AI, a $300M AI supercomputer at the University of Bristol.

Amazon launched a batch of its Project Kuiper internet satellites with the help of rival SpaceX’s Falcon 9 rocket.

Google agreed to pay more than $3B to source carbon-free hydropower from Brookfield Renewable Energy Partners.

SF-based newsletter platform Substack raised $100M in Series C financing, taking its valuation to $1.1B.

Elon Musk’s Neuralink labeled itself a “small disadvantaged business” in a federal filing before being valued at $9B in a funding round.  

Google officially set August 20 as the date for its Pixel launch event, where it will unveil its new Pixel smartphone lineup.

COMMUNITY

🎥 Join our next live workshop

Join our next workshop today at 4 PM EST with Dr. Alvaro Cintas, The Rundown’s AI professor. By the end of the workshop, you'll be able to use Perplexity Comet (and other alternatives) to automate your browsing experience.

RSVP here. Not a member? Join The Rundown University on a 14-day free trial.

See you soon,

Rowan, Jennifer, and Joey—The Rundown’s editorial team

Good morning, AI enthusiasts. OpenAI just gave ChatGPT its own computer — and the result is a benchmark-crushing agent that can actually handle real work.

With agent combining a variety of tools to autonomously browse, code, create presentations, and more within its own virtual workspace, the promise of 2025’s agentic revolution may have just taken a massive practical leap forward.

Reminder: Our next workshop is today at 4 PM EST — attend and learn how to use Perplexity Comet (and other alternatives) to automate your browsing experience. RSVP here.

In today’s AI rundown:

• OpenAI gives ChatGPT a computer

• Reflection AI’s Asimov agent for coding comprehension

• Automate your documentation workflow with Gemini CLI

• OpenAI beats all but one human in coding competition

• 4 new AI tools & 4 job opportunities

LATEST DEVELOPMENTS

OPENAI

🤖 OpenAI gives ChatGPT a computer

Image source: OpenAI

The Rundown: OpenAI just rolled out ChatGPT Agent, a major upgrade that allows the AI to control its own virtual computer to tackle complex workflows and agentic tasks, while setting new highs across benchmarks.

The details:

• Agent merges tools like Operator and Deep Research into a single system that can autonomously switch between browsing, coding, and document creation.

• OpenAI’s livestream showcased capabilities like booking travel, building presentations, shopping, creating a product, and setting up an order.

• Agent can also connect to apps like Gmail and GitHub, access APIs, and handle multiple tasks, permissions, and interruptions from the user.

• It shows SOTA performance across Humanity’s Last Exam (41.6%), Frontier Math, and a variety of real-world task benchmarks.

• OAI classified Agent as “high capability” for biological risks, enacting the strictest safety protocols, including live monitoring and user approvals.

Why it matters: OpenAI is following the path of agents like Manus, giving ChatGPT its own computer to drive a massive boost in agentic capabilities. While Operator was limited in real-world use cases at launch, Agent looks to combine all of AI’s strongest features — giving the first glimpse of where the agentic end game is heading.

See Rowan’s early access testing with ChatGPT Agent here.

TOGETHER WITH CONVEYOR

🛠️ AI was made for this

The Rundown: It's 2025 — security questionnaires should complete themselves, your customers should access your Trust Center instantly, and your content library shouldn't be a maintenance burden. Conveyor’s Customer Trust platform delivers exactly that with 95%+ accuracy.

Automate all your busywork so you can:

• Spend 90% less time answering security questionnaires and RFPs, even in tricky portals

• Skip knowledge maintenance — your content library stays current on its own

• Give your customers instant AI answers in your Trust Center

• Fully automate admin and collaboration tasks across teams and tools

Learn more and put Conveyor to work.

REFLECTION AI

⚙️ Reflection AI’s Asimov agent for coding comprehension

Image source: Reflection AI

The Rundown: Reflection AI, founded by ex-Google DeepMind researchers, just launched Asimov — an autonomous agent built to deeply understand codebases, business logic, and team knowledge that claims top performance over SOTA rivals.

The details:

• Asimov ingests not just code, but also architecture docs, emails, Slack threads, and project reports to build a persistent knowledge base for engineering teams.

• “Asimov Memories” let teams store and update tribal knowledge with natural language prompts, protected by role-based access controls.

• Asimov beat Claude Code with 82% developer preference in blind tests, using multiple “retriever” agents that feed findings to a central reasoning system.

• Reflection AI was founded by Misha Laskin and Ioannis Antonoglou, who previously worked on Gemini and AlphaGo at Google DeepMind.

Why it matters: Most coding assistants focus on generating code, but Reflection is betting that understanding existing codebases matters more than writing new ones. Asimov’s deeper dive and management of knowledge bases may be the start of AI agents as institutional memory — shifting how software is maintained and evolved.

AI TRAINING

⚙️ Automate your documentation workflow with Gemini CLI

The Rundown: In this tutorial, you will learn how to use Google's new command-line AI tool that automatically generates comprehensive README files and project documentation by analyzing your entire codebase for free.

Step-by-step:

1. Install it by typing npm install -g @google/gemini-cli in your terminal

2. Go to your desired folder via shell commands (“cd [path]”) and invoke the agent by typing gemini

3. Generate your README by asking: “Create a comprehensive README.md with installation, usage, and examples.”

4. Refine the documentation: “Add contributing guidelines and update the API documentation section.”

Pro tip: We just did a full workshop where you can see how to install, use, and automate workflows with Gemini CLI here.

PRESENTED BY INNOVATING WITH AI

💼 Does your business need an AI expert on call?

The Rundown: Innovating with AI has trained 1,000+ AI consultants, and now they’re launching their new directory of IWAI Certified AI experts – with exclusive early access for readers of The Rundown.

With this free directory, you can:

• Connect instantly with dozens of professional AI experts

• Get help with AI training, strategy, and implementation of workflows and agents

• Implement modern AI solutions with no fees, commission, or lock-in

Click here to request access to the IWAI Certified Professional directory.

AI & CODING

🥈 OpenAI beats all but one human in coding competition

Image source: Psyho (@FakePsyho on X)

The Rundown: OpenAI’s autonomous coding agent just placed second at the AtCoder World Tour Finals in Tokyo, with Polish (human) coder Psyho named the champion after a 10-hour showdown.

The details:

• Contestants faced complex optimization puzzles requiring them to guide digital robots through mazes while minimizing moves.

• Psyho clinched victory with a 9.5% margin after surviving on just 10 hours of sleep over three days, later posting "Humanity has prevailed (for now!).”

• The event marked the first time an AI model competed fully autonomously without human help against elite coders in a live programming final.

• Sakana AI tested their ALE-Agent alongside the official competition, achieving results that would have placed fifth overall.

Why it matters: OpenAI CEO Sam Altman previously said the company believed its models would be the best competitive programmers in the world by the end of the year, and this result looks to have them right on track. While Psyho took the crown, this is likely humanity's last gold medal in competitive programming over AI.

QUICK HITS

🛠️ Trending AI Tools

• 🤖 ChatGPT Agent - Let ChatGPT handle agentic tasks with its own computer

• 🎥 LTXV - Lighttrick's open-source video model with 60-second outputs

• 🖥️ Vision Desktop Share - Let Copilot view & analyze your screen in real-time

• ⚙️ Bedrock AgentCore - AWS’ suite of tools for deploying enterprise AI agents

💼 AI Job Opportunities

• 🧩 Pinecone - Customer Success Engineer

• 💼 Harvey - Head of Tax

• 🛠️ Cohere - Engineering Program Manager, North

• 📢 Luma AI - Account Executive, Advertising

📰 Everything else in AI today

Lovable founder Anton Osika announced a new $200M funding round that values the Swedish AI app-building startup at $1.8B.

Mistral rolled out major updates to its Le Chat platform, including Deep Research, Voice Mode, multilingual reasoning, Projects, and new image editing capabilities.

Hume AI released its EVI 3 speech-to-speech model via API, with the ability to clone voices and capture precise speaking styles for more emotion and personality.

Nvidia introduced Canary-Qwen-2.5B, a new SOTA speech recognition model that moved to the top spot on Hugging Face’s Open ASR leaderboard.

Suno released v4.5+, a new audio generation model with new song creation features including vocal swaps, playlist inspiration, and more.

Udio launched updates to its Styles feature for song generation, with new Blending, Library, and Artist Styles coming alongside expanded access for all users.

COMMUNITY

🎥 Join our next live workshop

Join our next workshop today at 4 PM EST with Dr. Alvaro Cintas, The Rundown’s AI professor. By the end of the workshop, you'll be able to use Perplexity Comet (and other alternatives) to automate your browsing experience.

RSVP here. Not a member? Join The Rundown University on a 14-day free trial.

See you soon,

Rowan, Joey, Zach, Alvaro, and Jason—The Rundown’s editorial team

Good morning, robotics enthusiasts. Shenzhen’s subway just got more crowded — with robots. Now, penguin-shaped delivery bots are cruising the city’s vast subway network, restocking snacks at 7-Elevens along the way.

Is this more than a convenience? Shenzhen is turning its subway into a robotic “middle mile” logistics network, offering a glimpse at the future of smart city supply chains.

In today’s robotics rundown:

• China’s delivery robots ride the subway

• Harvard creates creepy wormbot swarm

• Neura’s humanoids to build ships

• This jet-sized solar drone can fly for months

• Quick hits on other robotics news

LATEST DEVELOPMENTS

CHINESE ROBOTICS

🚊 China’s delivery robots ride the subway

Image source: Vanke

The Rundown: In Shenzhen, a fleet of 41 penguin-shaped delivery bots now zips through the subway stations and boards trains autonomously, ferrying drinks and snacks to more than 100 7-Eleven locations.

The details:

• The fleet of robots, each a meter tall with glowing LED “faces,” can navigate stations, elevators, and train carriages autonomously.

• VX Logistics, a Vanke subsidiary, manages the operation, with the fleet serving over 100 7‑Eleven stores located within the city’s vast subway network.

• Robots operate only during non‑peak hours to minimize congestion, replacing human couriers who otherwise need to park on the street and unload stock.

• Equipped with panoramic lidar and smart chassis design, the bots reportedly avoid collisions, handle subway platform gaps, and maneuver crowded spaces.

Why it matters: Shenzhen’s “Embodied Intelligent Robot Action Plan” has flipped the script on city logistics — robots now ride the rails, delivering straight to subway shops and skipping the curbside chaos. In the race for robotic dominance, this is a serious flex: In Shenzhen, even the subway has joined the robot workforce.

HARVARD UNIVERSITY

🪱 Harvard creates creepy wormbot swarm

Image source: Harry Tuazon/Georgia Institute of Technology

The Rundown: Harvard engineers just transformed the writhing, chaotic tangle of California blackworms into the blueprint for a new generation of robotic swarms — “roboworms” capable of autonomously entangling and moving as a unified collective.

The details:

• The team found inspiration after studying the collective behavior of California blackworms, which naturally entangle themselves to move and survive.

• The soft-bodied roboworms, about a foot long each, can physically latch onto and entwine with one another, forming dynamic multi-robot structures.

• Each roboworm contains onboard sensors and simple rules for local interaction, enabling decentralized decision-making, with no central controller.

• Their applications range from delicate search-and-rescue operations — where swarms could snake through rubble and debris — to planetary exploration.

Why it matters: The team aims to develop an untethered version that uses microfluidics to guide its action for full autonomy. Plus, the long-term goals of the project are to study the dynamics of group behaviors that emerge from entanglement, potentially creating supercharged AI systems that can scale walls and beyond.

NEURA ROBOTICS

🚢 Neura’s humanoids to build ships

Image source: Neura Robotics

The Rundown: Hyundai’s shipyards are getting a major robotic upgrade — not from its own robotics division or Boston Dynamics (which it owns), but from Germany’s Neura Robotics, whose humanoids will power next-generation robotic shipbuilders.

The details:

• Hyundai signed a deal with Neura Robotics to develop four-legged bots and specialized humanoids for shipbuilding, leveraging its humanoid platform 4NE1.

• HD Hyundai Samho will provide the hands-on testing environments at its shipbuilding facilities to get the robots up to speed.

• The 4NE1 robots are designed for advanced manipulation, adaptability, and teamwork, making them well-suited for complex assembly on the shipyard floor.

• Neura also recently announced a partnership with GFT Technology for its humanoids and raised $123M in Series B funding earlier this year.

Why it matters: This deal is a big shift in industrial automation, with one of the world’s top shipbuilders choosing cutting-edge, foreign-made humanoids over its own in-house solutions. The result of this team-up could push the arrival of truly intelligent, adaptable robots on some of the industry’s toughest front lines.

SKYDWELLER

☀️ This jet-sized solar drone can fly for months

Image source: Skydweller

The Rundown: U.S. startup Skydweller Aero joined forces with French defense giant Thales to develop a solar-powered autonomous drone that can stay aloft for months and has a 236-foot wide wingspan, even bigger than a Boeing 747.

The details:

• Dubbed Skydweller, the drone’s surface is covered with 17K high-efficiency solar cells, making it capable of flying for months without refueling.

• It can cruise at high altitudes, making it ideal for persistent intelligence, surveillance, and reconnaissance (ISR) missions.

• Onboard AI analyzes massive data streams mid-flight, autonomously filtering information and deciding what is to be transmitted to command centers.

• Skydweller is also featherweight compared to a jumbo jet, topping out at just 2.5 metric tons, making it 160 times lighter than a fully loaded Boeing 747.

Why it matters: This giant AI-powered drone could patrol the world’s oceans nonstop — silently gathering data day and night, streaming real-time maritime intel, and never burning a drop of fuel or needing to land. If Skydweller hits its targeted milestones for endurance flight, it could rewrite the playbook for airborne surveillance.

QUICK HITS

📰 Everything else in robotics today

Hugging Face said that it hit $1M in sales for its Reachy Mini desktop robot just five days after opening up orders.

Ex-Waymo engineers launched Bedrock Robotics with $80M in funding to retrofit existing construction equipment with 24/7 autonomous systems.

Gartner predicts that 1 in 20 supply chain managers will manage robots, not humans, by 2030.

Baidu and Uber plan to place thousands of Baidu’s Apollo Go autonomous vehicles on the Uber platform in markets outside the U.S. and mainland China.

China’s robot wolves have reportedly been deployed in a human-drone drill for the Chinese People’s Liberation Army.

“Rizzbot,” a Unitree humanoid, has been seen wearing a cowboy hat and Pride flags while walking through traffic and engaging pedestrians in West Hollywood, LA.

Tesla sharply expanded its geofenced operating area for its invite-only Robotaxi pilot in Austin, less than a month after it debuted the service.

MIT researchers developed a deep-learning control system that teaches soft, bio-inspired robots to move and follow commands from just a single image.

Chinese flying car company XPeng Aeroht nabbed $250M in Series B funding to accelerate its Land Aircraft Carrier modular flying car.

COMMUNITY

🎥 Join our next live workshop

Join our next workshop this Friday, July 18th, at 4 PM EST with Dr. Alvaro Cintas, The Rundown’s AI professor. By the end of the workshop, you'll be able to use Perplexity Comet (and other alternatives) to automate your browsing experience.

RSVP here. Not a member? Join The Rundown University on a 14-day free trial.

See you soon,

Rowan, Jennifer, and Joey—The Rundown’s editorial team

Good morning, AI enthusiasts. The past year of AI video acceleration has brought hyper-realism, smooth motion, and perfect lip-syncs — all trapped in frustrating 5-second bursts.

But an open-weight model just shifted the conversation from fragmented edits to real-time directing of 60-second generations — with LTXV’s update handing creators the keys to a brand-new style of AI video production.

In today’s AI rundown:

• LTXV unlocks 60-second AI videos

• New ChatGPT agents for Excel, PowerPoint

• Research market conditions and create financial dashboards

• Self-driving AI lab discovers materials 10x faster

• 4 new AI tools & 4 job opportunities

LATEST DEVELOPMENTS

LIGHTRICKS

🎥 LTXV unlocks 60-second AI videos

Image source: Lightricks

The Rundown: Lightricks just released an update to its open-weights LTXV model, now allowing for image-to-video generations over 60 seconds long — streamed in real time, with live prompt control and efficient performance on consumer GPUs.

The details:

• The model streams video live as it generates, returning the first second instantly while building scenes continuously without cuts.

• Users can apply control inputs throughout generation, adjusting poses, depth, and style mid-stream for dynamic scene evolution.

• LTXV is trained on fully licensed data, with direct integration with LTX Studio’s production suite and the ability to run efficiently on consumer devices.

• The open-source model has both 13B and mobile-friendly 2B parameter versions, available free on GitHub and Hugging Face.

Why it matters: With most video models topping out at just a few seconds of output, LTXV’s real-time, 60-second outputs change what’s possible with current AI video content. The shift from waiting minutes for short clips to directing longer scenes as they generate will open up entirely new creative workflows.

TOGETHER WITH ARTISAN

🔎 Identify and prospect local businesses with an AI BDR

The Rundown: Struggling to identify local business decision makers? Our AI BDR, Ava, taps into a database of 200M+ local Google businesses and does fully autonomous outreach, so you can focus on closing deals, not chasing leads.

Ava operates within Artisan, which consolidates every tool you need for outbound:

• 300M+ high-quality B2B prospects

• Automated lead enrichment across 10+ data sources

• Multi-channel outreach with full email deliverability management

• Live intent signal scraping, website visitor ID, fundraise announcements & more

Book a demo and supercharge your sales team.

OPENAI

📊 New ChatGPT agents for Excel, PowerPoint

Image source: Ideogram / The Rundown

The Rundown: OpenAI is reportedly developing ChatGPT agents that can create and edit spreadsheets or presentations directly in chat and bypass the need for suites like Microsoft Office and Google Workspace, according to a report from The Information.

The details:

• ChatGPT will feature dedicated buttons below the search bar to generate spreadsheets and presentations using natural language prompts.

• The outputted reports will be directly compatible with Microsoft’s open-source formats, allowing users to open them across common applications.

• An early tester reported “slow and buggy” performance from the ChatGPT agents, with a single task taking up to half an hour.

• OpenAI reportedly also has a collaboration tool allowing multiple users to work together within ChatGPT, but there is no information on its release yet.

Why it matters: The ability to create sheets and presentations in ChatGPT, away from Microsoft’s Office empire, feels like a personal shot — but OpenAI “steamrolling” some of the most-used business apps isn’t a surprising move. Still, “slow and buggy” won’t cut it if it truly wants to displace Microsoft and Google’s widely adopted products.

AI TRAINING

📊 Research market conditions and create financial dashboards

The Rundown: In this tutorial, you will learn how to use Grok 4's advanced research and canvas to conduct comprehensive market analysis and build dynamic financial forecasting models.

Step-by-step:

1. Access Grok and choose Grok 4 in the model dropdown menu

2. Request comprehensive market research for [INDUSTRY] financial forecasting by analyzing current economic conditions.

3. Ask for scenario planning: “Create three financial scenarios: best case, most likely, and worst case for the next 18 months.”

4. Create visualizations using its Canvas by asking it to code a visual financial dashboard with revenue projections and key economic indicators.

Pro tip: Grok 4's real-time access keeps your forecasts current as market conditions evolve.

PRESENTED BY SLACK FROM SALESFORCE

🔎 Find everything with enterprise search in Slack

The Rundown: Slack just introduced enterprise search, turning a single search bar into your company’s knowledge engine to pull answers across your data and from third-party apps without a single context switch.

In this 20-minute webinar, you’ll learn:

• How Slack unifies structured and unstructured data by connecting to apps like Salesforce, Google Drive, Microsoft Teams, and more

• Best practices for structuring questions to get the best results from enterprise search

• How enterprise search with Agentforce equips employees to ramp faster & work smarter

Learn more about enterprise search and watch the webinar.

AI RESEARCH

🧪 Self-driving AI lab discovers materials 10x faster

Image source: North Carolina State University

The Rundown: North Carolina State University researchers developed an AI-powered, self-driving lab that continuously streams chemical experiments, collecting 10 times more data than traditional systems and speeding up the search for new materials.

The details:

• The new system uses dynamic, real-time experiments instead of waiting for each chemical reaction to finish, keeping the lab running continuously.

• By capturing data every half-second, the lab’s machine-learning algorithms quickly pinpoint the most promising material candidates.

• The approach also significantly cuts down on the amount of chemicals needed and slashes waste, making research more sustainable.

• Researchers said the results are a step closer to material discovery for “clean energy, new electronics, or sustainable chemicals in days instead of years”.

Why it matters: Many of the scientific breakthroughs coming in the AI boom have been medical-related, and materials science underpins many of the fields set to drive these world-changing innovations. Just like we’ve seen across other domains, the theme remains the same — AI taking complex research from years to days.

QUICK HITS

🛠️ Trending AI Tools

• 🧠 Semrush AI Toolkit - Benchmark your brand’s AI visibility vs. competitors, monitor sentiment, and uncover high-impact AI prompts by real customers*

• 🎬 Act-Two - Runway’s next-generation AI motion capture model

• ⏺️ Record Mode - Capture & summarize meetings & voice notes with ChatGPT

• ☎️ AI Mode - Use Google’s AI agents to make calls on your behalf

*Sponsored Listing

💼 AI Job Opportunities

• 🏗️ Glean - Solutions Architect

• 📊 Abridge - Product Manager, Enterprise

• 🚀 OpenAI - Design Execution Manager, Stargate

• 🎬 Luma AI - Account Executive, Entertainment

📰 Everything else in AI today

Meta reportedly poached Jason Wei and Hyung Won Chung from OpenAI, with the two researchers previously contributing to both the o1 model and Deep Research.

Anthropic is gaining Claude Code developers Cat Wu and Boris Cherny back, with the duo returning after joining Cursor-maker Anysphere earlier this month.

Microsoft is rolling out Desktop Share for Copilot Vision to Windows Insiders, allowing the app to view and analyze content directly on users’ desktops in real-time.

Scale AI is laying off 14% of its staff in a restructuring following the departure of CEO Alexandr Wang and other employees as part of a multibillion-dollar investment by Meta.

OpenAI is reportedly creating a checkout system within ChatGPT for users to complete purchases, with the company receiving a commission from sales.

Anthropic is receiving interest from investors for a new funding round at a valuation of over $100B, according to a report from The Information.

AWS unveiled Bedrock AgentCore in preview, a new enterprise platform of tools and services for deploying AI agents at scale.

COMMUNITY

🎥 Join our next live workshop

Join our next workshop this Friday, July 18th, at 4 PM EST with Dr. Alvaro Cintas, The Rundown’s AI professor. By the end of the workshop, you'll be able to use Perplexity Comet (and other alternatives) to automate your browsing experience.

RSVP here. Not a member? Join The Rundown University on a 14-day free trial.

See you soon,

Rowan, Joey, Zach, Alvaro, and Jason—The Rundown’s editorial team