Senior AI Journalist
AI Drug Reverses Biological Age Markers in Early Trials
In a groundbreaking development, AI-driven pharmaceutical company Insilico Medicine has announced that one of its AI-designed drug candidates, rentosertib, shows promising signs of reversing markers associated with biological aging. This significant news, reported by The Decoder, highlights the increasing impact of artificial intelligence in accelerating complex scientific research and drug discovery.
Rentosertib was initially developed for patients suffering from idiopathic pulmonary fibrosis (IPF), a severe condition that leads to lung tissue scarring. During a trial involving 42 patients, the drug demonstrated improvements in lung function. Crucially, blood samples collected during this trial were later analyzed, forming the basis of a new study published in Nature Biotechnology.
The core of this new analysis involved applying six independent aging clocks—AI models designed to predict biological age from blood proteins—to the patient data. These clocks, developed by various independent research teams at institutions like Harvard, Oxford, and Beijing, consistently indicated a lower biological age for patients treated with rentosertib compared to those who received a placebo. The most pronounced effect observed was a reduction of three to four years in biological age by week four, with one clock even suggesting a six-year reduction.
It is important to note that these findings suggest a shift in blood protein patterns that the AI models interpret as a younger biological age, rather than a direct, measurable reversal of aging in patients. However, Nobel laureate in chemistry Michael Levitt emphasized the significance of the consistent agreement across diverse models, stating that the agreement across independent models is what matters most.
The blood values might only look younger because the lungs work better and the body is under less overall strain. But the dose that helped the lungs most differed from the one that cut predicted biological age the most, suggesting an effect at least partly independent of lung function. The researchers also compared treated patients’ blood proteins against more than 55,000 profiles from the UK Biobank, a database that tracks how certain proteins typically change with age. According to the company, rentosertib reversed exactly those changes.
Experts like cardiologist Eric Topol and Harvard Medical School’s Vadim Gladyshev acknowledge the encouraging nature of the study, calling it the first study that shows biological age can be reduced. However, they also stress the need for larger trials, particularly in healthy individuals, to confirm these effects and understand the broader implications beyond IPF patients.
Insilico Medicine’s innovative approach involves two AI systems: one identifies disease-relevant proteins from vast health data and scientific literature, and the other analyzes protein structures to design matching molecules. This method allowed Insilico to pinpoint TNIK as a target for both aging and pulmonary fibrosis, accelerating the drug development process significantly. The rapid progress from target identification to drug candidate in about 18 months, with rentosertib now in Phase III trials, showcases the transformative potential of AI in biopharmaceutical research. Pharma giants like Eli Lilly have already recognized this potential, investing in Insilico Medicine to co-develop AI-driven drugs.
Source: The Decoder
Kalanick’s Atoms Ventures into Robotaxi Territory with Uber Collaboration
Travis Kalanick, the co-founder of Uber, appears to be steering his latest venture, Atoms, into the burgeoning robotaxi industry. A recent report in the Financial Times, corroborated by TechCrunch, indicates that Atoms is gearing up for a significant hiring drive and potential acquisitions aimed at establishing itself as a major contender in the autonomous vehicle sector.
This strategic pivot follows Atoms’ substantial $1.7 billion funding round earlier this summer, spearheaded by Andreessen Horowitz. Despite the considerable investment, Kalanick had remained largely tight-lipped about the company’s precise objectives until now. The Financial Times report sheds light on these ambitions, revealing that Atoms has engaged in discussions with Uber regarding the potential integration of Atoms’ robotaxi technology into Uber’s expansive ride-hailing network. Notably, Uber has already invested $100 million in Atoms, a detail previously confirmed by TechCrunch.
While sources close to the matter emphasize that robotaxis are not the sole focus of Atoms’ long-term strategy, this direction aligns with Kalanick’s characterization of the recent funding as unfinished business. The move also resonates with Atoms’ earlier acquisition of Pronto, an autonomous mining startup that was previously led by Anthony Levandowski, Uber’s former self-driving chief.
The robotaxi market is a fiercely competitive landscape, with numerous companies vying for dominance. Atoms’ entry, backed by significant capital and Kalanick’s industry experience, could reshape the competitive dynamics. The potential for a symbiotic relationship with Uber, a global leader in ride-sharing, offers Atoms a unique advantage in deploying and scaling its autonomous vehicle technology.
The development of safe and reliable autonomous driving systems is a complex endeavor, requiring substantial investment in research, development, and infrastructure. Atoms’ reported plans for a hiring spree and acquisitions suggest a concerted effort to build a robust technical and operational foundation. The company’s focus on this area could also drive further innovation within the broader autonomous vehicle ecosystem, potentially leading to more advanced and efficient self-driving solutions.
This expansion into robotaxis also underscores the broader trend of technology giants and well-funded startups investing heavily in AI and robotics to transform various industries. As autonomous technology matures, its applications are expected to diversify, extending beyond personal transportation to logistics, public transit, and specialized services. For companies looking to power their own AI and robotics initiatives, reliable and scalable infrastructure is paramount. A robust Contabo VPS can provide the necessary computational power and flexibility to support such ambitious projects.
Source: TechCrunch
The Uncanny Valley of AI-Generated Menus: When Food Looks Too Perfect
Generative AI is transforming various aspects of design and content creation, but its foray into restaurant menus has revealed an unexpected challenge: the sameness problem. As reported by TechCrunch, AI-generated food images often possess an unsettling, almost alien perfection that triggers an uncanny valley effect in human observers, leading to discomfort and unease rather than appetite.
The issue stems from how these AI models are trained. Large Language Models and diffusion models, which power AI image generators, learn from vast datasets of existing imagery. When tasked with creating a menu, they tend to draw from a corpus of commercially optimized food photography, such as advertisements from major fast-food chains. These commercial images are already highly stylized and aim for an idealized, often unrealistic, aesthetic. The AI amplifies these characteristics, producing visuals that are symmetrically flawless, unnaturally smooth, and devoid of the subtle imperfections that make real food appealing.
Alex Lisle, CTO of Reality Defender, succinctly describes this phenomenon as almost like an alien trying to make a pizza without understanding its core principles. This convergence in AI-generated outputs, while not as severe as model collapse, still degrades the quality and authenticity of the visuals. The models essentially reinforce an already narrow aesthetic, leading to homogenized and predictable designs that lack originality and warmth.
Furthermore, an experiment by a user named Labtec demonstrated that repeated edits to an AI-generated menu image can exacerbate the problem. With each iteration, the food items become progressively more rounded and smooth. This highlights a significant pitfall for restaurants attempting to use AI for menu design: rather than enhancing visual appeal, repeated refinement can push the images further into the realm of the artificial and unappetizing.
Lee Rainie, director of the Imagining the Digital Future Center at Elon University, points out that humans possess an innate, almost subconscious ability to discern AI-generated content from genuine articles. Research from the University of Duisburg-Essen in Germany supports this, finding that AI-generated food images that closely resemble reality can evoke greater disgust than obviously fake ones, illustrating the uncanny valley effect in action.
The implications extend beyond just food. The digital age is blurring the lines between reality and artificiality, challenging fundamental assumptions about visual evidence. As AI-generated content becomes more prevalent and sophisticated, the ability to distinguish authentic visuals from synthetic ones will become an increasingly critical skill. For now, in the culinary world, the lesson is clear: for food to truly tempt, it needs a touch of genuine imperfection that AI, in its pursuit of flawless symmetry, struggles to replicate.
Source: TechCrunch
What to Read Next
- OpenAI Keeps Re-Enabling Training on Your Data: How to Protect Your Privacy in 2026
- Hallucination: What It Means in AI and Why It Matters (2026 Guide)
- OpenAI Launches GPT-Live-1 API with Full-Duplex Speech: A Quantum Leap in Real-Time Conversational AI
- How to Run Stable Diffusion on a VPS: An Affordable Guide for AI Enthusiasts
- Browse all AI Stack Digest articles
Bookmark aistackdigest.com for daily AI tools, reviews, and workflow guides.
This article was produced with the assistance of AI tools and reviewed by the AIStackDigest editorial team.