AI security risks are shifting from models to workflows after malicious extensions stole chat data from 900,000 users & prompt injections abused AI to
Get the latest international news and world events from around the world.
Google now lets you change your @gmail.com address, rolling out
Google has confirmed that it’s now possible to change your @gmail.com address. This means that if your current email is [email protected], you can now change it to [email protected].
Gootloader now uses 1,000-part ZIP archives for stealthy delivery
The Gootloader malware, typically used for initial access, is now using a malformed ZIP archive designed to evade detection by concatenating up to 1,000 archives.
In doing so, the malware, which is an archived JScript file, causes many tools to crash when trying to analyze it.
According to researchers, the malicious file is successfully unpacked using the default utility in Windows, but tools relying on 7-Zip and WinRAR fail.
View a PDF of the paper titled A Disproof of Large Language Model Consciousness: The Necessity of Continual Learning for Consciousness, by Erik Hoel
Scientific theories of consciousness should be falsifiable and non-trivial. Recent research has given us formal tools to analyze these requirements of falsifiability and non-triviality for theories of consciousness. Surprisingly, many contemporary theories of consciousness fail to pass this bar, including theories based on causal structure but also (as I demonstrate) theories based on function. Herein I show these requirements of falsifiability and non-triviality especially constrain the potential consciousness of contemporary Large Language Models (LLMs) because of their proximity to systems that are equivalent to LLMs in terms of input/output function; yet, for these functionally equivalent systems, there cannot be any falsifiable and non-trivial theory of consciousness that judges them conscious. This forms the basis of a disproof of contemporary LLM consciousness. I then show a positive result, which is that theories of consciousness based on (or requiring) continual learning do satisfy the stringent formal constraints for a theory of consciousness in humans. Intriguingly, this work supports a hypothesis: If continual learning is linked to consciousness in humans, the current limitations of LLMs (which do not continually learn) are intimately tied to their lack of consciousness.
New synaptic formation in adolescence challenges conventional views of brain development
Researchers from Kyushu University discovered a previously unrecognized synaptic “hotspot” that forms during adolescence, challenging the long-held view that adolescent brain development was dominated by synaptic pruning. This hotspot fails to form in mice carrying a schizophrenia-associated gene, pointing to a potential link between adolescent synaptic formation and psychiatric disorders, including schizophrenia.
Adolescence marks an important transition not just socially and physically, but neurologically. During this period, higher cognitive functions such as planning, problem-solving, and decision-making gradually mature. Yet, the underlying mechanisms of neural circuit development remain poorly understood.
Key to this process are synapses—the functional connections between neurons allow information to flow through the brain. Previously, it has long been hypothesized that synapse numbers increase during childhood and then decrease during adolescence. It has also been proposed that excessive “synaptic pruning,” a process that refines neural circuits by eliminating unused or weak connections, may lead to neuropsychiatric disorders. One example is schizophrenia, a condition characterized by hallucinations, delusions, or disorganized thinking.
What Is Nanotechnology? The Atomic Future Waiting to Begin
The idea never died, progress is still being made.
Nanotechnology was once imagined as the next great technological revolution—atom-by-atom manufacturing, machines as small as cells, and materials we can only dream of today. Instead, it stalled. While AI, robotics, and nuclear surged ahead, nanotech faded into the background, reduced to buzzwords and sci-fi aesthetics.
But the idea never died.
We can manipulate matter at the atomic scale. We can design perfect materials. We can build molecular machines. What’s been missing isn’t physics—it’s ambition, investment, and the will to push beyond today’s tools.
In this interview with futurist J. Storrs Hall, we explore what nanotechnology really is, why it drifted off course, and why its future may finally be on the horizon. If AI was a “blue-sky fantasy” until suddenly it wasn’t, what happens when someone decides nanotech deserves the same surge of talent, money, and imagination?
Neocortical Chandelier Cells Developmentally Shape Axonal Arbors through Reorganization but Establish Subcellular Synapse Specificity without Refinement
Cortical chandelier cells (ChCs) are one of the most distinct and uniform IN subtypes (Howard et al., 2005; Woodruff et al., 2010). ChC axons exhibit a characteristic geometry with many prominent vertical branches, whose terminals are specialized into strings of synaptic varicosities (cartridges) directly apposed to AISs of PNs (Jones, 1975; Szentagothai, 1975; Somogyi, 1977). Because the AIS is the site of action potential initiation, ChCs can have decisive control over spike generation in a PN ensemble, thereby regulating synchrony and oscillation of network activity (Klausberger et al., 2003; Szabadics et al., 2006; Glickfeld et al., 2009). The striking stereotypy and specificity of this axonal and synaptic organization make ChCs an ideal system to study basic cellular events of IN wiring, such as axonal branching and subcellular synapse targeting. Another advantage is that individual AISs, which can be labeled with AIS-specific markers such as anti-AnkyrinG (AnkG) antibodies, can be unambiguously identified because they are spatially separable from neighboring AISs (Jenkins and Bennett, 2001; Taniguchi et al., 2013).
Study:
Basically, the study is carried out carefully and provides novel insights in the development of neocortical chandelier cells. While the manuscript is well written; both reviewers suggest that authors get help with linguistic editing.
The reviewers also agree that the interpretation that the non-synaptic varicosities may represent early stage branch points is somewhat surprising, considering their abundance at an age at which very little branching occurs. The reviewers ask the authors to expand their discussion considering the following arguments/ideas.
In Fig. 3E, it is shown that on P 28, i.e. close to adulthood and after pruning and remodeling, there are still around 40% off-target varicosities, which, if the single example in Fig. 5B is true and representative, would mean that 40% of all varicosities are non-functional! While it is possible that such varicosities could be the origins of branch points to remodel the axon, it is hard to imagine that ChC remodel 40% of their axon at any given time in adulthood. Please illustrate in a revised manuscript better resolved and uncolored original EM data to really show that these large varicosities do not form any synapses. Furthermore, it would be somewhat surprising that their numbers do not decrease after p21 when most axonal growth is expected to end. What about the possibility that these varicosities are in fact very early stage synaptic boutons that fail to mature/stabilize because the appropriate postsynaptic target is missing? Considering that inhibitory boutons (formed by other interneuron subsets) continue to be formed and lost into adulthood seem to make this option also quite likely.