BioWire Weekly - 031
BioTech News
Happy Friday evening, Readers. I hope you were relentless this week.
In keeping with our new format, we’re closing out this round of BioWire Bytes (technically, two weeks’ worth) with a recap of what we’ve covered. Think of this as a navigation tool and synopsis, ideal for those who haven’t had the time to read each post or who just prefer scanning the highlights. Personally, I think it pairs best with a Saturday morning cup of coffee.
The topics this past week:
AlphaGenome Predicts Gene Regulation
ChatBots Aren’t Really Thinking.
Neuralink Study Makes Big Strides.
Tesla Launches Robotaxis
The Robots are Coming! Amazon deploys its 1 millionth robot.
Grok 4 Advances in Humanities Last Exam.
First, if you enjoy these updates, consider subscribing and becoming a part of our growing community!
BioWire Bytes 001 - AlphaGenome Predicts Gene Regulation
To all of my fellow Americans, Happy Independence Day! We are living through the most important period in human history, at the epicenter of many significant advancements and innovations in biotechnology, health, and artificial intelligence. It is the framework of our country and the diverse people who inhabit it that made this progress possible. While …
AlphaGenome is DeepMind’s newest genomic AI; it’s a single, powerful model that reads huge stretches of DNA and predicts how genes are regulated inside cells. It already has shown capabilities in a real-world disease prediction, pinpointing how an overlooked DNA mutation switches a cancer gene on. The availability of an open API and may enable AlphaGenome to quickly become a new standard tool for geneticists.
BioWire Bytes 002 - ChatBots Aren't Actually Thinking
The Illusion of Thinking in AI Reasoning Models
Large language models (LLMs) like ChatGPT and Claude still don't genuinely "think"; they primarily excel at pattern matching but struggle with true logical reasoning. Newer reasoning-augmented models, designed with techniques like chain-of-thought prompting, handle moderate puzzle complexities better but collapse when puzzles become too intricate, indicating fundamental limits to their "thinking" abilities. Ultimately, this study emphasizes that even advanced reasoning models simulate rather than genuinely perform logical thought, challenging how we evaluate true reasoning in AI.
BioWire Bytes 003 - Neuralink Clinical Study Makes Big Strides
Neuralink Announces 7 Total Patients Implanted with Brain-Computer Interface in Summer Update
Neuralink's first human clinical trial, "Telepathy," now includes seven participants across multiple countries, already transforming their lives by enabling high-speed, precise control of digital devices and even robotic limbs through thought alone. Demonstrations showed participants effortlessly gaming online, using robotic arms, and performing previously impossible tasks, emphasizing the practical and life-changing nature of this technology. With ambitious plans to scale implants into thousands of channels and expand into vision restoration and cognitive therapies, Neuralink seems on the brink of reshaping what human-computer interaction means.
BioWire Bytes 004 - Tesla Launches Robotaxi
Tesla’s long-awaited RoboTaxi, well, at least a version of it, has finally launched. It’s exactly what it sounds like: a self-driving taxi summoned through a ride-hailing app, like Uber or Lyft. While that’s undeniably exciting, there are some caveats: this is a very limited rollout, and it’s not yet the CyberCab model we’ve all been waiting for.
Tesla has officially launched its first robotaxi pilot in Austin, using a small fleet of Model Y vehicles fully operated by vision-based AI, no driver, steering wheel, or pedals required. Unlike competitors relying on lidar and radar, Tesla's fully integrated, camera-only approach allows ordinary production cars to become autonomous through software updates, significantly lowering cost and potentially enabling rapid scalability. This pilot sets the stage for Tesla’s ambitious future, including the dedicated Cybercab vehicle, and signals a looming race with Waymo for robotaxi dominance.
BioWire Bytes 005 - The Robots are Coming! Amazon deploys 1 millionth robot.
If it seems like our coverage lately has leaned heavily into AI, robotics, and automation, you're not imagining things. There are just too many important milestones occurring to overlook, and we quietly passed another. Just last month, Amazon announced that its robotic workforce had crossed the one-million mark. These robots come in all shapes and sizes…
Amazon recently hit a staggering milestone by deploying its one-millionth robot across over 300 fulfillment centers globally, enhancing delivery speed and warehouse efficiency, especially during high-demand events like Prime Day. Powering this robotic fleet is DeepFleet, Amazon’s generative AI that acts like a traffic controller to optimize robot movements, significantly cutting down congestion and improving overall efficiency. Interestingly, rather than displacing workers, this surge in robotics has driven Amazon to upskill and hire more human employees in advanced technical roles.
BioWire Bytes 006 - Grok 4 Advances in Humanities Last Exam
Another day, and another major AI development. This time, we’re looking at xAI’s Grok 4, the latest large language model (LLM) to be released. Is it just me, or would anyone else welcome a decade-long pause in further AI development? We’ve barely begun to understand the implications and real-world use cases of the current generation of LLMs, yet we keep…
xAI's latest large language model, Grok 4, has dramatically raised the bar for AI reasoning, significantly outperforming competitors on the notoriously challenging "Humanities Last Exam" benchmark, especially when using its unique multi-agent "heavy" mode. Unlike previous models, Grok 4 can collaborate internally, employing multiple reasoning agents and external tools simultaneously to tackle complex problems previously thought unsolvable by AI. While skeptics question whether these benchmarks truly measure "thinking," Grok 4's capabilities hint at a rapidly approaching future where AI fundamentally reshapes multiple fields.
These newsletters take significant effort to put together and are totally for the reader's benefit. If you find these explorations valuable, there are multiple ways to show your support:
Engage: Like or comment on posts to join the conversation.
Subscribe: Never miss an update by subscribing to the Substack.
Share: Help spread the word by sharing posts with friends directly or on social media.