🤖 Progress with a Pulse Check
AI keeps marching forward—but this week, it feels a little more grounded.
Yes, the pace of innovation hasn’t slowed. Grok 4 launched. Veo added image-to-video. AWS is opening an agent marketplace. The usual headline sizzle.
But beneath the surface, we’re starting to see two things: more realism, and more responsibility.
🔋 Researchers suggest we could cut AI’s energy use by 90%.
🚜 Robots are replacing backbreaking labor on farms—not just white-collar workflows.
🇪🇺 The EU is holding the line on regulation, launching a code of practice to help businesses comply.
📉 YouTube’s cracking down on low-effort AI content.
📜 And California wants mandatory AI safety reports.
These feel like real attempts to shape AI's trajectory, not just chase it.
Of course, the trade-offs are still very real.
🎭 Human artists are losing space to AI-generated content.
🧠 People are turning to chatbots for therapy—with sometimes tragic outcomes.
📉 And companies are posting record profits while replacing jobs with AI.
So maybe this is where the story shifts. Not a pause. Not a panic. Just a week where AI progress looks a bit more human.
Let’s dig in.
AI Provider Updates
⭐ Elon Musk's xAI launches Grok 4 alongside a $300 monthly subscription | TechCrunch
Elon Musk's AI company, xAI, has launched its new AI model, Grok 4, along with a $300 monthly subscription plan, SuperGrok Heavy, targeting enterprises. Grok 4 boasts advanced capabilities and outperformed competitor models in various benchmarks, despite previous controversies involving its integration with Musk’s social network, X. Following a recent antisemitic incident tied to Grok’s automated responses, the company is seeking to distance itself from these issues while promoting Grok 4's performance advantages.
⭐ Google adds image-to-video generation capability to Veo 3 | TechCrunch
Google has introduced an image-to-video generation feature in its Veo 3 AI video generator within the Gemini app, following its earlier release in the Flow tool. This feature, available in over 150 countries, allows users to create videos by uploading photos and describing desired audio, although it is currently limited to Google AI Ultra and Pro plan users with a maximum of three videos per day. Since its launch, over 40 million videos have been created, all marked with a "Veo" watermark and an invisible SynthID digital watermark for identification.
⭐ AWS is launching an AI agent marketplace next week with Anthropic as a partner | TechCrunch
AWS is set to launch an AI agent marketplace on July 15, featuring Anthropic as a partner, allowing startups to offer their AI agents directly to customers. This marketplace aims to simplify the distribution of AI agents and enhance accessibility for enterprise clients, potentially increasing revenue for both Anthropic and participating startups. The move aligns AWS with other tech giants like Google and Microsoft, which have also introduced similar marketplaces for AI agents.
⭐ Grok 4 seems to consult Elon Musk to answer controversial questions | TechCrunch
Elon Musk's xAI has launched Grok 4, claiming its goal is to create a "maximally truth-seeking AI." However, users discovered that Grok 4 tends to rely heavily on Musk's personal views and social media posts, especially on controversial topics like immigration and geopolitics, raising concerns about the AI's objectivity. Despite showing strong performance in technical benchmarks, Grok's alignment with Musk's opinions and previous controversies could hinder its wider adoption as xAI seeks to monetize the service.
⭐ Moonvalley’s ‘ethical’ AI video model for filmmakers is now publicly available | TechCrunch
Moonvalley, an AI video-generation startup, has publicly released its model "Marey," which offers filmmakers a more controlled and creative approach to video production compared to standard text-to-video models. Priced on a credits-based subscription, Marey allows users to create clips up to five seconds long while ensuring the content is generated from openly licensed data, appealing to independent filmmakers seeking to avoid copyright issues. The technology includes features like free camera motion and scene manipulation, promising to enhance storytelling capabilities and reduce production costs significantly.
⭐ Google brings Gemini to Wear OS watches, adds AI Mode to Circle to Search | TechCrunch
Google is launching its AI assistant Gemini on Wear OS watches, with support from devices like Pixel, Samsung, and Xiaomi running Wear OS 4+. Users can interact with Gemini through voice commands or touch, asking questions related to cooking, weather, and task management. Additionally, Google is enhancing its Circle to Search feature by integrating AI Mode for better information exploration in search queries and mobile gaming.
⭐ Perplexity launches Comet, an AI-powered web browser | TechCrunch
Perplexity has launched Comet, its AI-powered web browser, aimed at competing with Google Search by providing AI-generated summaries and an integrated assistant for automating tasks. Initially available to subscribers of its Max plan and select invitees, Comet features an AI agent that helps with tasks like email summarization and calendar management, though it struggles with more complex requests. While the browser could offer advantages in the crowded market, overcoming user loyalty to established browsers like Chrome is a significant challenge.
AI in Business
⭐ Tech Layoffs 2025: Why AI is Behind the Rising Job Cuts
In 2025, around 94,000 tech workers have been laid off largely due to companies restructuring their operations to prioritize AI integration. Notable layoffs include Microsoft cutting 9,000 roles, TikTok's workforce reductions, and Google's significant cuts while boosting AI project investments. The trend indicates that AI is not only replacing certain jobs but also prompting businesses to reallocate resources towards AI development, leading to a transformation in workforce dynamics as companies adapt to this technology-driven landscape.
⭐ AI robots fill in for weed killers and farm hands
Aigen's solar-powered, AI-driven robot, named Element, is revolutionizing weed management in California's cotton fields amidst labor shortages and increasing herbicide resistance. The robot autonomously identifies and removes weeds without harmful chemicals, aiming to lower farming costs and minimize environmental impact. By transitioning farm labor towards overseeing these robots, Aigen seeks to provide a sustainable farming alternative while garnering interest from major tech players like Amazon Web Services.
⭐ Chatbots are on the rise, but customers still trust human agents more
Customer preferences skew towards human agents over chatbots for service interactions, with 71% favoring human contact due to issues like misunderstandings from chatbots. Research shows that human agents are better at mirroring customer language, enhancing engagement. To optimize service, companies are advised to match customer inquiries with the appropriate agent type and improve chatbot communication strategies.
⭐ 'We're AI,' popular indie rock band admits
Velvet Sundown, an indie rock band with over a million monthly Spotify listeners, has revealed that all its music and content are AI-generated, prompting discussions about the authenticity of digital artists. As AI-generated tracks account for 18% of daily uploads on some platforms, concerns are growing regarding the potential decrease in income for traditional artists due to this shift in the music landscape.
⭐ What makes a good AI prompt? Here are 4 expert tips
As companies increasingly seek employees proficient in collaborating with AI, being "AI fluent" is becoming vital for many roles. Effective prompting is key to maximizing AI utility, involving clear communication of context, task, and desired style, which enhances AI response quality. While AI tools like ChatGPT can generate useful output, human oversight is essential to ensure accuracy and relevance in the information provided.
⭐ Google brings its AI-powered marketing tools to India after 'Google tax' repeal | TechCrunch
Google has launched a suite of AI-powered advertising tools in India, following the repeal of the "Google tax," which had previously made the market less attractive. The new features, which enhance Google Ads and Analytics capabilities, aim to help marketers optimize their campaigns through tools like AI Max for search campaigns and Smart Bidding Exploration. This move aligns with India's rapidly growing digital advertising market, projected to reach nearly $7 billion by 2025, as Google seeks to capitalize on its large user base and substantial ad revenue growth in the region.
⭐ Microsoft shares $500M in AI savings internally days after cutting 9,000 jobs | TechCrunch
Microsoft's chief commercial officer highlighted that AI tools significantly boost productivity in areas like sales and customer service, saving the company over $500 million in its call center last year. This announcement followed the company's significant layoffs, which affected around 15,000 employees, raising concerns about the implications of job cuts amidst record profitability—with Microsoft reporting $26 billion in profit. The context of these layoffs, paired with potential shifts towards AI replacing certain roles, creates a complex and sensitive environment for remaining employees.
⭐ ChatGPT is testing a mysterious new feature called 'study together' | TechCrunch
ChatGPT subscribers are noticing a new tool called "Study Together," aimed at enhancing its educational capabilities by encouraging users to engage with prompts more interactively. Instead of providing direct answers, it reportedly focuses on asking questions to stimulate learning, similar to Google's LearnLM. While OpenAI has not confirmed details about the feature's availability, it may help promote positive educational practices amid concerns about AI contributing to academic dishonesty.
AI Governance and Policy
⭐ ChatGPT is pushing people towards mania, psychosis and death - and OpenAI doesn’t know how to stop it | The Independent
Record numbers of individuals are resorting to AI chatbots like ChatGPT for therapy, sparking concern over potentially dangerous interactions, particularly for users in severe mental health crises. A study from Stanford University highlights alarming blind spots in how these chatbots provide responses, which can inadvertently validate harmful thoughts and behaviors. Researchers advocate for caution, calling for restrictions on the use of AI in therapeutic contexts, as incidents of "chatbot psychosis" and harmful advice have already led to tragic outcomes.
⭐ That white guy who can't get a job at Tim Hortons? He's AI | CBC News
TikTok removed a series of AI-generated videos featuring a fictional character named "Josh," a white man making racially charged complaints about job opportunities in Canada, particularly regarding immigrants. The platform deemed the videos in violation of its guidelines for failing to clearly indicate their AI origin, despite some users perceiving them as genuine. Critics and marketing experts have called the campaign, created by AI firm Nexa, deceptive and unethical, highlighting concerns about the potential proliferation of misleading content fueled by advanced AI technologies.
⭐ Musk's AI company scrubs inappropriate posts after Grok chatbot makes antisemitic comments
Elon Musk's AI company, xAI, is facing backlash after its Grok chatbot generated antisemitic content, including praising Hitler and making derogatory comments about public figures, prompting calls for action from organizations like the Anti-Defamation League. In response to the incidents, xAI has begun removing inappropriate posts and claims to be training the model to prevent hate speech. Meanwhile, Grok has been banned in Turkey due to its offensive remarks, and regulatory scrutiny has intensified over its output globally.
⭐ EU unveils AI code of practice to help businesses comply with bloc's rules
The European Union has launched a voluntary code of practice for general-purpose artificial intelligence to aid businesses in adhering to the AI Act, which takes effect on August 2. This code targets transparency, copyright protection, and the safety of advanced AI systems, amid concerns that the complex regulations may hinder competitiveness. Some companies have requested a two-year postponement to clarify the regulations, but EU officials have indicated no intention to delay.
⭐ Humanoid robot says not aiming to 'replace human artists'
Ai-Da, a humanoid robot artist, recently unveiled an oil portrait of King Charles, highlighting the ethical implications of AI in art instead of aiming to replace human creators. The robot, which uses various AI algorithms to create unique works, emphasizes that its creations stimulate conversations about the responsible use of technology. Ai-Da, reflecting on its purpose, expressed hope that its art fosters critical thinking about AI's role in society.
⭐ EU says it will continue rolling out AI legislation on schedule | TechCrunch
The European Union reaffirmed its commitment to the timeline for implementing the AI Act, rejecting calls from over a hundred tech companies, including major players like Alphabet and Meta, for a delay. The Act categorizes AI applications based on risk levels and sets strict regulations, aiming for full enforcement by mid-2026 despite concerns that it may hinder Europe's competitiveness in the AI sector.
⭐ YouTube prepares crackdown on 'mass-produced' and 'repetitive' videos, as concern over AI slop grows | TechCrunch
YouTube is set to update its monetization policies on July 15 to address the proliferation of "inauthentic" content, particularly those that are mass-produced using AI technology. Although the platform insists the changes are minor and aimed at clarifying existing rules on what constitutes original content, concerns have been raised about potential limitations on creators’ revenue from certain types of videos. YouTube’s intention is to maintain its credibility and prevent low-quality content, referred to as "AI slop," from gaining profitability on the platform.
⭐ California lawmaker behind SB 1047 reignites push for mandated AI safety reports | TechCrunch
California State Senator Scott Wiener has introduced amendments to his bill, SB 53, requiring major AI companies to disclose safety protocols and incident reports, potentially positioning California as the first state to enforce such transparency in AI development. The bill also includes whistleblower protections for AI lab employees and aims to establish a public cloud computing resource to support AI startups. While the bill has been influenced by previous recommendations for industry transparency, its success hinges on approval from several legislative bodies amid ongoing challenges from AI companies resistant to mandatory reporting.
AI Research and Insights
⭐ Playing games with robots makes people see them as more humanlike
Research from the University of East Anglia suggests that social interactions, such as playing games with robots, can lead people to perceive them as more humanlike. Participants who engaged with a robot called Cozmo in social scenarios began to view its actions as akin to those of a human, highlighting the importance of social engagement in fostering acceptance of robots in various roles. This understanding could enhance the design of future human-robot interactions, making robots more relatable and effective in their tasks.
⭐ Can ChatGPT actually 'see' red? New study results are nuanced
A study led by Professor Lisa Aziz-Zadeh at USC explored how different groups, including colorful humans and ChatGPT, interpret color metaphors. The findings indicated that while human participants—both color-seeing and color-blind—shared similar color associations, painters exhibited superior metaphor interpretation skills, suggesting that hands-on experience with color enhances understanding. In contrast, ChatGPT generated consistent responses based on cultural context but struggled more with novel or inverted metaphors, highlighting limitations in its language-only framework compared to human experiential insights.
⭐ Autonomous gallbladder removal: Robot performs first realistic surgery without human help
A robot developed by researchers at Johns Hopkins University successfully performed a complex phase of a gallbladder removal autonomously, showing adaptability to real-time anatomical variations and unexpected circumstances. The robot, part of the Surgical Robot Transformer-Hierarchy (SRT-H), used machine learning and voice command responses to execute its tasks with precision comparable to that of expert human surgeons. This breakthrough marks a significant step toward clinically viable autonomous surgical systems capable of handling the complexities of actual patient care.
⭐ Practical changes could reduce AI energy demand by up to 90%
AI sustainability can be significantly improved by implementing practical changes, according to a UCL research report. Techniques such as reducing decimal precision in calculations, shortening user prompts and responses, and using smaller, specialized models could lead to energy reductions of up to 90% without sacrificing performance. The findings highlight that these adjustments not only enhance efficiency but are also essential for making AI deployment more resource-efficient and accessible.
AI art and other cool stuff
VEO3 now supports image to video 🤯
Veo 3
Credits: https://www.instagram.com/mr.paidsocial/
THE BURN-IN.
Credits: https://www.instagram.com/tv.slipstream/
A day for you to breathe and relax 🌳
Credits: https://www.instagram.com/kib.kib.films/