Deepmind CEO Demis Hassabis compares the arrival of AGI to ten times the industrial revolution in a tenth of the time. "I sometimes quantify AGI as 10 times the industrial revolution at 10 times the speed. So unfolding over a decade instead of a century," Hassabis says in the 20VC podcast. He sees a "very good chance of it being within the next 5 years," an assessment that hasn't changed since 2010, when co-founder Shane Legg predicted it would take 20 years: "I think we're pretty much on track."
Getting there still requires several major advances, including continuous learning, long-term planning, better memory architectures, and greater consistency. Hassabis describes current systems as "jagged intelligences," "really amazing at certain things when you pose the question in a certain way, but if you pose a question in a slightly different way they can still fail at quite elementary things." Scaling continues to deliver results, "although they're a bit less than they were at the start of all of this scaling."
Hassabis also points to a growing perception gap. "Today and in the next year things are a bit overhyped in AI," he says. But looking further out, "it's still very underappreciated how revolutionary this is going to be in the time scale of about 10 years."
Alibaba's Qwen team built HopChain to fix how AI vision models fall apart during multi-step reasoning
When AI models reason about images, small perceptual errors compound across multiple steps and produce wrong answers. Alibaba’s HopChain framework tackles this by generating multi-stage image questions that break complex problems into linked individual steps, forcing models to verify each visual detail before drawing conclusions. The approach improves 20 out of 24 benchmarks.
AI safety research firm Lyptus Research has published a new study on the offensive cybersecurity capabilities of AI models. The study is based on the METR time-horizon method and involved testing with ten professional security experts.
According to the findings, AI's offensive cyber capability has been doubling every 9.8 months since 2019, and since 2024, that pace has accelerated to every 5.7 months. Opus 4.6 and GPT-5.3 Codex can now solve tasks at a 50 percent success rate with a two-million-token budget that would take human experts roughly three hours to complete.
Offensive cyber capability of AI models over time: From GPT-2 (2019) to Opus 4.6 and GPT-5.3 Codex (2026), the time horizon grew from 30 seconds to roughly three hours. The doubling time accelerated from 9.8 months (since 2019) to 5.7 months (since 2024). | Image: Lyptus Research
Performance jumps significantly with higher token budgets: GPT-5.3 Codex goes from a 3.1-hour to a 10.5-hour time horizon when given ten million tokens instead of two million. The researchers say this suggests they're still underestimating the actual rate of progress. Open-source models trail their closed-source counterparts by about 5.7 months.
Alibaba's Qwen team makes AI models think deeper with new algorithm
Reinforcement learning hits a wall with reasoning models because every token gets the same reward. A new algorithm from Alibaba’s Qwen team fixes this by weighting each step based on how much it shapes what comes next, doubling the length of thought processes in the process.
