Oct 16, 2026 · 7 min listen · Last updated October 16, 2026
From storyflo. This is your daily audio brief. It's Theo. October 16th, tech roundup — five stories, here's number one. Let's get into it. First, from TechCrunch. Even Benchmark thinks AI is eroding the venture pact against backing rivals. So, there’s this interesting shift happening in venture capital, especially around AI.
Listen · storyflo · tech
Daily Tech Brief · October 16th
0:00-6:31
Pick your daily storyteller
Subscribe to match with Theo, Jessica, Chloe, Mason, Brock — your voice, every brief.
Audio pre-rendered by Storyflo · cached + delivered from the edge
Even Benchmark thinks AI is eroding the venture pact against backing rivals
So, there’s this interesting shift happening in venture capital, especially around AI. Traditionally, firms like Benchmark have avoided backing competing companies, but they’re starting to rethink that stance. At a recent TechCrunch Disrupt event, Benchmark partners admitted that the rapid evolution in AI is straining those old rules. For instance, two AI companies, OpenAI and Anthropic, share many of the same investors, which is a departure from the norm.
A recent incident with a startup called Factory highlights this tension. They fired an adviser for allegedly discussing a job with Cognition, a direct rival, while being involved in Factory's decisions. This situation has drawn attention from prominent investors, including Vinod Khosla, who defended the adviser and criticized Factory. Benchmark’s partners, like Eric Vishria, acknowledged that as companies pivot quickly, conflicts are becoming inevitable, even if they prefer to maintain exclusivity at the early stages.
Chetan Puttagunta pointed out that as AI lowers the cost of software development, founders are becoming more ambitious and eager to claim their market space, leading to more overlaps. Peter Fenton described this issue as “kryptonite” for investors, emphasizing the importance of transparency between founders and their backers. While Benchmark isn’t throwing out the old rules entirely, their candid acknowledgment of these conflicts might make founders reconsider how they view loyalty in their partnerships. It’s a fascinating moment in the industry, for sure.
SaaSpocalypse revisited: How AI is breaking software's old economics
Zoho, a major player in the SaaS world, is feeling the pinch from rising AI costs and a shifting market. Co-founder Sridhar Vembu recently announced a freeze on pay increases and hiring for their 19,000 employees, citing skyrocketing server costs and a tough geopolitical landscape. This reflects a broader trend in the software industry, where traditional companies are scrambling to adapt as AI makes software cheaper to build but raises operational expenses.
Interestingly, initiatives like ArtCraft are emerging, showcasing how a single developer can create alternatives to established products using AI, which could disrupt pricing structures. Enterprises are also considering building their own software rather than purchasing it, driven by AI's capabilities. McKinsey's survey highlights that a significant portion of organizations are opting to develop in-house solutions, indicating a shift in how software is sourced.
As AI continues to evolve, it’s not just about creating software but also about maintaining and securing it. While traditional firms have advantages, like brand loyalty and customer relationships, they might struggle to keep their pricing power intact as customers explore cheaper, AI-generated options. The landscape is changing, and it’s going to be fascinating to see how these dynamics play out in the coming years.
Claude now leads 26% of Anthropic’s R&D, Meta drops an 816B open model & Gemini 4 Argon
Hey! So, it’s been a wild week in AI, right? Meta is really stepping up its game with Muse Wonder, which is now ranked #2 on Arena Text, just behind Google’s Gemini 4 Argon. They also dropped Muse Spark Open, an 816 billion parameter model that’s making waves. It’s fascinating how they’re positioning it—Meta’s calling it the most capable open-weight model in the U.S. right now. The pricing strategy is interesting too; they’re keeping the top-tier models behind a paywall, which feels like a smart move.
Then there’s Google’s Gemini 4 Argon, which is back with a bang. It’s already topping the charts on multiple evals and is being praised for its training run. The demos showed some impressive capabilities, especially in generating creative outputs. I loved hearing about how Geneviève Huskens created a racing game using Argon; it’s a neat example of how these models can be used in unexpected ways.
On the Anthropic front, Claude is now leading 26% of their R&D efforts, and they’ve tightened up their security measures, which is causing some hiccups for users. The tech is getting more sophisticated, but it sounds like it’s also becoming a bit more restrictive. It’s a balancing act, right? Overall, there’s a lot happening, and it feels like we’re just scratching the surface of what these models can do. Can’t wait to see how it all unfolds!
The AI Companion Question (Robert Wright & Paul Bloom)
Note: Paid NZN subscribers can now get a 50 percent discount on a one-year subscription to Paul’s excellent Substack newsletter, Small Potatoes. The discount coupon is at the bottom of this newsletter, below the paywall. This episode includes an Overtime segment that’s available to paid subscribers. If you’re not a paid subscriber, we of course encourage you to become one. If you are a paid subscriber, you can access the complete conversation either via the audio and video on this post or by setting up the paid-subscriber podcast feed, which includes all exclusive audio content.
Computers Made of Human Brain Tissue Identify Braille Letters In Breakthrough
Scientists have made a breakthrough toward the mind-bending goal of creating computers made of brains by demonstrating that human neural tissue can classify letters of the Braille alphabet, according to a study published on Friday in Frontiers in Artificial Intelligence. A team transmitted electrical signals corresponding to Braille letters to human brain organoids, which are living neural tissues derived from stem cells.
A version of Muse Spark Open was released today, fulfilling their original promise but with a model weaker than their Spark 1.3. It’s also worth watching the Tesla Roadster demo - not strictly AI but nothing cooler than real life flying cars. Our title story today is a rare one on interpretability and alignment.
The triple-OS NexPhone runs Android, Linux, and Windows 11, but you’ll have to wait until 2027 for it
Usually, you're only stuck with one OS on your phone, but what if you could pick between Android, Linux, and Windows 11? It sounds ambitious, but that's a major selling point for NexPhone, which has been in the works for 14 years now. The phone was originally scheduled to release later this year, but unfortunately, it has been pushed back a few months.
Tesla’s New Roadster Can Somehow Do Corkscrews in the Air (Without Exploding—Yet)
The new Tesla Roadster can’t quite fly, but it can accelerate to 60 miles per hour in under a second using boosters, drive upside down, and propel itself through corkscrews in the air—at least under the conditions of the prototype’s unveiling at a SpaceX test site in McGregor, Texas, on Thursday night. “We need an audience to prove it’s not AI or SI,” Tesla and SpaceX CEO Elon Musk told the audience (referring to the White House’s preferred language, “Super Intelligence”).
On the markets — Kalshi traders have been actively repricing this story in the last day.
Gemini 4 Argon is now rolling out to Google’s Antigravity for AI Pro subscribers
Google is now rolling out support for Gemini 4 Argon to Antigravity, its AI coding tool. Gemini 4 Argon is Google’s latest flagship AI model, announced at the end of September and having widely rolled out to users earlier this week. Now, it’s also available within Antigravity. Antigravity is an agentic development platform, as Google describes it, which leverages AI and agents to assist in coding and building apps.
TypeSafe’s Jev ‘even more susceptible to attack’ than LLMs, says security vendor
Although the Jev model rolled out last month by TypeSafe claims to have significant advantages over traditional LLMs, a security report published Wednesday argued that its streamlined approach makes it more vulnerable to some kinds of attack, compared to traditional LLMs. “In some cases, [Jev] is even more vulnerable to attacks, with both adversarial text attacks and prompt injection being able to disrupt the model’s outputs, resulting in incorrect and potentially devastating actions,” said the blog post from researchers at HiddenLayer .