Best of LinkedIn: ICT & Tech Insights CW 21/ 22

00:00:00: This episode is provided by Thomas Allgaier and Freeness, based on the most relevant LinkedIn posts about ICT and tech insights from CW-TwentyOne and TwentyTwo.

00:00:09: Freenes supports ICT enterprises in the form of delivering precise ICT market and pricing intelligence that analyst subscriptions and existing databases cannot provide – you can find more info.

00:00:21: Imagine a custom microchip, right?

00:00:24: Like no larger than the tip of an needle powering a pacemaker inside a patient's chest.

00:00:28: And now imagine that same microscopic chip is actually actively repelling a cryptographing attack from a multi-million dollar quantum supercomputer

00:00:37: which sounds like sci-fi.

00:00:38: but welcome to The Deep Dive!

00:00:39: Our mission today is parsing the absolute top tech and ICT trends we've seen dominating professional networks like LinkedIn over last two weeks

00:00:47: Right...and we are zeroing in on infrastructure levels shifts targeting those of you who are actually building and defending the enterprise.

00:00:54: No fluff today?

00:00:55: Yeah, we're completely bypassing this standard hype cycle.

00:00:57: if you work in tech You know that noise right now is just deafening.

00:01:01: It

00:01:01: really is.

00:01:02: But

00:01:03: when look closely at what engineers or strategists are debating a stark picture emerges We're tracking a multi-billion dollar bedrock shift in quantum architecture The countdown clock on post quantum cryptography, which is terrifying by the way.

00:01:19: Terrifying yeah and How autonomous AI agents are basically forcing us to rip out our entire philosophy of network defense?

00:01:29: Well

00:01:29: let's start with the loudest conversation in the room right now Which is definitely quantum computing.

00:01:34: I mean it has officially graduated from a theoretical lab experiment into a massive geopolitical and economic battleground.

00:01:42: The sheer scale of the financial commitments being made right now is just altering landscape, so Steve McDowell, Naveen N., and Ilysis Caruso they all kind of dissected this recent ten billion dollar commitment from IBM to deliver fault tolerant quantum computing by twenty-twenty nine.

00:01:59: okay i have to pause you right there because we've been hearing that quantum quote unquote five years away for the last twenty years.

00:02:06: Oh

00:02:06: yeah, The Perpetual Five Years

00:02:08: It's exactly!

00:02:18: R&D announcements.

00:02:19: Well, the

00:02:19: distinction really lies in The Target which is their Starling system.

00:02:23: So IBM isn't just trying to increase the raw number of physical quibits anymore.

00:02:27: They are targeting two hundred

00:02:30: logical quibitz

00:02:31: and they're backing this up by building America's first pure play quantum chip foundry.

00:02:37: Like Ulysses Caruso pointed out, this is a foundational infrastructure play.

00:02:41: The organizations that control the bedrock of quantum and AI infrastructure.

00:02:45: they'll effectively dictate terms for next era computing.

00:02:49: Right but before we dig into hardware We really need to clarify that term For anyone listening who isn't currently living in a physics lab.

00:02:56: A logical quibbit Is very different from physical one.

00:02:58: Vastly different

00:02:59: because Quantum states are incredibly fragile like a slight temperature variance or a stray microwave photon, and the quibbit just loses its state which is called decoherence.

00:03:09: Right

00:03:09: it just falls apart!

00:03:10: Exactly so.

00:03:12: to get one reliable error-free logical quibit you typically have to network thousands of physical quibits together just to handle the error correction.

00:03:22: Which Is Why?

00:03:23: Alex Aparov's analogy was so sharp.

00:03:25: he compared the current state of quantum computing to The aviation industry at the nineteen tenths.

00:03:29: Oh,

00:03:30: I love that comparison.

00:03:31: Right because we know flight is possible But we don't know what the dominant plane is gonna look like.

00:03:35: yet We have superconducting quibits trapped ion systems photonic processors neutral

00:03:41: atoms

00:03:42: exactly.

00:03:42: Everyone is trying to solve That exact error correction bottleneck but using completely different physics

00:03:48: and we actually saw two specific hardware breakthroughs in the last two weeks that attack this fragility from fascinating angles.

00:03:55: First, Gene Lee and Pradumna Gupta broke down Microsoft's new Majurana II chip.

00:04:00: Oh the topological one?

00:04:01: Yeah so Microsoft isn't trying to build a better error correction algorithm.

00:04:04: they're trying to bypass the problem entirely using topological quibits

00:04:09: which is I mean the physics behind topological quibits are notoriously difficult to engineer, but the concept is brilliant.

00:04:16: Instead of storing quantum information in a single localized particle that can get disrupted easily they distribute the information across the entire structure of an engineered nanowire.

00:04:26: Right.

00:04:27: think about it like tying a knot into shoelace.

00:04:29: If you drop the shoelaces or even step on them The knot's still there.

00:04:37: not a single point.

00:04:38: That's a great way to visualize it!

00:04:39: Right,

00:04:40: so if Microsoft can scale this...it theoretically makes quibits a thousand times more reliable without that massive overhead of thousands of physical quibbits just for error correction.

00:04:50: But then you have the flip side-a completely different approach highlighted by Ethan Batraski.

00:04:55: He pointed out this profound systems milestone from a company called Adam Computing.

00:05:00: they demonstrated a system that survived failing hardware through continuous mid-computation quibbit replenishment.

00:05:07: Which is basically the ship of Theseus applied to a quantum state, because previously if a physical quibit died in the middle of complex calculation The whole thing just crashed.

00:05:16: you had start over

00:05:17: right

00:05:18: but atom computing proved You can dynamically inject fresh quibits into the array To replace the dead ones without stopping the calculation.

00:05:27: I mean They are making the architecture behave like continuously operating infrastructure.

00:05:32: And with hardware evolving this fast, you know the natural question becomes what can it actually do in the real world today?

00:05:39: Because enterprise leaders don't buy potential right they buy capabilities

00:05:43: They want ROI.

00:05:44: Exactly so.

00:05:45: Sanjay Vishwakarma and Lara Jehi highlighted this milestone project involving IBM The Cleveland Clinic and Reichen .They successfully simulated a trypsin protein consisting of twelve thousand six hundred and thirty five atoms.

00:05:58: Wow And to put that in perspective for you, simulating molecular behavior is exponentially difficult.

00:06:04: Like just six months prior the limit of this specific method was around three hundred atoms.

00:06:09: Yeah

00:06:10: it's a massive jump.

00:06:11: Going into twelve thousandth is roughly forty x increase in size.

00:06:15: But mechanism behind it really matters here.

00:06:19: It wasn't purely a quantum calculation, it was a hybrid approach.

00:06:23: Right!

00:06:23: They paired ninety-four quantum quibits with two of the world's most powerful classical supercomputers.

00:06:29: The Classical Machines handled the bulk of standard computational heavy lifting while the Quantum Processor specifically tasked with complex quantum mechanical interactions of the molecule.

00:06:40: It kind of proves that quantum centric super computing where they divide labor is realistic path to commercial utility especially in things like drug discovery.

00:06:50: But,

00:06:50: you know classical computing algorithms aren't just rolling over yet.

00:06:54: NucreB shared this incredible reality check a recent claim of quantum supremacy?

00:06:59: Oh I saw that

00:07:01: right where complex problem was supposedly impossible for classical computers to solve.

00:07:05: well it's quietly refuted.

00:07:07: physicists adapted the mathematical algorithm from the nineteen eighties compressed the tensor network data and literally solved The Impossible Problem on A Standard Laptop.

00:07:17: So standard laptop.

00:07:18: That is a great reminder that software optimization still matters immensely.

00:07:23: And Greg Dick brought up another structural reality check, the real bottleneck holding back.

00:07:28: that hybrid classical quantum future isn't just raw quip at power it's

00:07:37: massive temporal mismatch.

00:07:39: A top-tier GPU operates in nanoseconds, but depending on the architecture some quantum gates take microseconds to execute

00:07:47: right.

00:07:47: they're too slow.

00:07:48: yeah The gpu is constantly just sitting around waiting for the quantum processor To finish its thought.

00:07:53: so until engineers solve that interface latency?

00:07:55: The hybrid model's gonna hit a ceiling.

00:07:57: But if we assume those engineering hurdles will be cleared and the capital flow strongly suggests They Will Be We really have to confront the darker side of this exponential curve Which brings us to the cryptographic cliff we are rapidly approaching.

00:08:10: You're talking about Q-Day?

00:08:11: Yeah, Q-day!

00:08:12: So Alex Pruden released a stark report warning that day quantum computers can break today's standard RSA and elliptic curve.

00:08:19: cryptography could arrive as early as twenty thirty to twenty thirty three

00:08:23: which is tomorrow in enterprise terms

00:08:25: literally tomorrow.

00:08:26: Standard public key encryption relies on math problems like prime factorization.

00:08:32: That would take classical computer thousands of years.

00:08:35: But a mature quantum computer running Shor's algorithm, it could solve them in minutes.

00:08:40: Trillions of dollars and digital assets, interbank transfers, encrypted health records... It all relies on the assumption of computational difficulty that is just evaporating!

00:08:50: And migrating an enterprise to post-quantum cryptography or PQC Is easily a decade long project Absolutely.

00:08:58: but here's this structural problem with pqc The new algorithms require significantly more computational overhead.

00:09:04: Like, it's one thing to push a cryptographic update into the massive rack of servers in a data center.

00:09:09: But what about the edge?

00:09:10: What about tiny industrial sensors on a pipeline or the pacemaker we mentioned at very start of deep dive?

00:09:16: You cannot run heavy post-quantum algorithms without draining it instantly!

00:09:21: Which is exactly why Sunny Freedy's post was so critical... Researchers at MIT have developed a custom microchip specifically designed to run post-quantum cryptography on wireless biomedical devices.

00:09:33: Oh wow!

00:09:34: Yeah, they engineered the silicon to be hyper efficient meaning we can secure things like insulin pumps and ingestible biosensors against future quantum attacks without sacrificing their operational lifespan.

00:09:46: That's incredible.

00:09:47: And on the pure defense side of things, David Steendhoek shared an absolutely wild breakthrough from ETH Zurich that redefines how we generate the cryptographic keys themselves.

00:09:56: they successfully generated mathematically certified perfect random numbers

00:10:02: because standard computer randomness isn't actually random right

00:10:05: exactly.

00:10:05: it relies on a pseudo-random number generator which is really just an algorithm.

00:10:10: So if a well-funded attacker like a nation state with a quantum computer knows the algorithm and the seed value they can reverse engineer The pattern and break the key right.

00:10:18: so to solve this each Zurich use quantum entanglement.

00:10:22: They fired microwave photons across a thirty meter vacuum tube Entangled two quivits an extracted pure quantum uncertainty

00:10:29: that is wild

00:10:30: in the thirty meter distance.

00:10:31: Is the crucial mechanism here?

00:10:33: They had to measure the quibits faster than light could travel down a tube.

00:10:37: Since nothing exceeds the speed of light, it mathematically proved that chips couldn't be secretly signaling each other.

00:10:43: The resulting numbers were proven flawlessly physically random.

00:10:47: I mean That is foundation for truly unhackable cryptography right there.

00:10:51: But while scientific community building these ultimate defenses Enterprise strategy lagging dangerously behind Dr Katrina R highlighted terrifying strategic oversight.

00:11:03: What's that?

00:11:04: Well, the industry is heavily focused on the threat of Harvest Now decrypt later.

00:11:08: Which Is The Idea?

00:11:09: That Attackers Are Horting Encrypted Data Today To Decrypt It In Twenty Thirty.

00:11:13: But She Warns We Are Entirely Ignoring Harvest NOW Forged Later.

00:11:16: Ah!

00:11:17: Because If A Quantum Computer Breaks The Underline Cryptography it Doesn't Just Read Your Secrets...It Breaks Public Key Infrastructure.

00:11:25: The Digital Certificates That Prove a Website, A Software Update Or A Digital Identity Is Actually Authentic Exactly.

00:11:32: If trust itself becomes forageable and attacker can impersonate your bank, CEO or operating systems update server you wouldn't even be able to trust the software patch that you download.

00:11:44: fix vulnerability in first place

00:11:46: which is exactly why experts like Garfield Jones and authors of new book post quantum leadership are emphasizing this concept of crypto agility.

00:11:57: Enterprise architecture can no longer hard code cryptographic standards into their applications.

00:12:02: Your systems must be modular, built to swap out compromise algorithms dynamically without tearing down the entire

00:12:08: infrastructure.".

00:12:09: But the reality is most enterprise networks are nowhere near that level of agility.

00:12:13: and while quantum computing is the existential threat of tomorrow The timeline for artificial intelligence is right.

00:12:19: now we are entering AI network supercycle.

00:12:22: Oh yeah Helen Yu shared data from the keynote at Cisco Live.

00:12:26: that should honestly terrify network administrators.

00:12:43: And the telemetry data proves it.

00:12:45: Every AI agent generates roughly four hundred and fifty percent more network traffic than a human executing the exact same workflow,

00:12:53: because

00:12:54: they're constantly polling databases, executing API calls validating data...and don't sleep!

00:13:00: Most legacy enterprise networks simply do not have bandwidth or architecture to handle that swarming

00:13:05: behavior.".

00:13:06: And this traffic explosion is forcing an immediate shift toward action-based zero trust model… Because traditional Zero Trust asks, who are you and what role do you have?

00:13:15: But action-based Zero Trust asked What specific actions are executing right now at machine speed?

00:13:21: And does it deviate from your behavioral baseline.

00:13:23: Right because if a compromise AI agent gets loose inside of perimeter It can inflict years worth damage in matter hours.

00:13:29: Think about this way.

00:13:31: A human threat actor is like master safecracker.

00:13:35: They meticulously look for one perfect flaw.

00:13:39: But an AI agent swarm isn't a safecracker, it's like unleashing billion microscopic nanobots to test every single atom of the vault door simultaneously until structural weakness yields.

00:13:51: That is terrifying image!

00:13:52: And Rafe Pilling reported on this exact scenario.

00:13:55: A threat actor built an AI-assisted lab using models like Claude to iteratively write test and refined malware specifically engineered to evade top-tier endpoint detection platforms from CrowdStrike, and Microsoft.

00:14:08: And Annette Nima pointed out that this dramatically increases the volume in the complexity of malicious code way beyond human capacity.

00:14:15: security analysts simply cannot manually triage the output of an automated AI malware factory.

00:14:21: I am gonna challenge the premise that AI is creating a new cybersecurity problem though Michael Spotz offered a really brilliant counter perspective.

00:14:28: He argues that AI is simply exposing a very old, very ignored problem.

00:14:32: for decades enterprises have just limped along with messy active directories unresolved technical debt and sprawling identity access.

00:14:39: he has

00:14:39: absolutely right.

00:14:40: Enterprises relied on the buffer of human limitation.

00:14:44: attackers moved relatively slowly but now The attackers are operating at machine speed while enterprise security teams are still operating at well committee speed

00:14:53: Exactly.

00:14:54: You cannot schedule a change advisory board meeting to approve a firewall

00:14:57: rule

00:14:58: when an AI agent is iterating attack vectors hundreds of times per minute.

00:15:03: The disparity in operational velocity,

00:15:07: which means the enterprise operating model just has to evolve.

00:15:10: Arthur S shared Gartner research, stressing that organizations must adopt an assume breach mentality.

00:15:15: Assume Breach prevention is failing.

00:15:17: the architectural goal has to be survivability and rapid cyber resilience.

00:15:21: And we saw a perfect example of why committee speed as fatal.

00:15:24: just last week on public Astro issued an urgent warning about a CVSS.

00:15:28: nine point eight vulnerability in Windows servers.

00:15:31: That's CVE.

00:15:31: twenty-twenty six forty N eighty nine at a nine point ten basically means The House Is Actively On Fire.

00:15:38: Oh, it's a nightmare flaw in the NetLogOn protocol.

00:15:41: It requires absolutely no credentials No user interaction meaning nobody even has to click a phishing link...it is wormable.

00:15:48: So it can self-propagate across a network autonomously and it grants The attacker system level access which Is the highest possible

00:15:56: privilege.

00:15:57: Literally one malformed Network packet And the attacker owns the domain controller.

00:16:03: They control the keys To entire enterprise kingdom.

00:16:06: If your organization requires a week of committee meetings to test and deploy that patch, you're compromised before the meeting even gets scheduled.

00:16:13: And this widespread failure of voluntary security hygiene is finally triggering regulatory crackdowns.

00:16:19: Salima Al-Bardak noted that secure by design, it's no longer just a best practice.

00:16:23: It has become a strict legal liability.

00:16:25: Regulations like the UK's PSTI Act and EU Cyber Resilience Act are fundamentally changing market.

00:16:31: Right Device Security mandatory lifecycle patching and coordinated vulnerability disclosures They're all becoming prerequisites for market access

00:16:38: Which perfectly bridges into our final theme today.

00:16:40: Because we can engineer the most elegant top logical quibits.

00:16:44: We can mandate perfect quantum randomness and deploy autonomous AI defenses, but if the human beings operating The Enterprise aren't aligned with the technology...the entire digital transformation just

00:16:55: fails.".

00:16:56: Yeah George Culley provided a pretty sobering reality check regarding massive enterprise resource planning or ERP migrations like moving legacy systems to SAP cloud.

00:17:07: He pointed out that these multi-million dollar projects rarely fail because the IT department can't figure how to install software.

00:17:14: Right They fail, because organizational transformation is infinitely harder than technical implementation.

00:17:20: You cant write a script To fix a stubborn spreadsheet culture.

00:17:24: Software does not fix broken business processes or siloed departments.

00:17:27: It just digitizes the dysfunction.

00:17:29: Digitizes

00:17:30: the disfunction, I love

00:17:31: that!

00:17:31: And Manuel Bergen expanded on this arguing the transformation fails when human resources and the employee experience are treated as an afterthought.

00:17:39: Right, if you deploy an advanced AI-driven software suite but fail to train the workforce or align the tool with their actual daily workflows.

00:17:48: You aren't transforming the business...you're just creating what he calls

00:17:53: dumb change.

00:17:54: Although Sabrina Stork did offer a note of optimism pointing out that modern migration assistance are finally reducing the technical friction of these legacy transitions which makes them achievable for midsize enterprises.

00:18:06: But

00:18:06: The Human Element still remains the anchor of both transformation and security.

00:18:11: Roberto Ishmael Panino wrote a phenomenal piece reframing cybersecurity entirely, he argues it is not a technical hurdle It isn't an act of care

00:18:20: An Act Of Care.

00:18:21: Yeah good security hygiene Isn't just about protecting the corporate database.

00:18:24: Its' about protecting your colleagues Your clients And your digital community.

00:18:29: And Matthias Mullert takes that philosophy and really backs it up with science in his new book, Beautiful Security.

00:18:35: He uses concepts like chaos theory fractals an evolutionary biology to prove that security doesn't have To be a reactive exhausting cycle of panic and patching.

00:18:45: when It's engineered With the human operator in mind it can actually Be a manageable understandable and deeply human endeavor.

00:18:51: we've covered immense ground today From the physics of tying quantum knots and nanowires to the sheer speed of AI swarms exposing legacy corporate governance.

00:19:01: As you look at the infrastructure, you manage.

00:19:03: consider this final thought technology is evolving an exponential machine-speed curve but The true bottleneck in your enterprise isn't that coherence time of a logical quibbit And it isn't the bandwidth of your network.

00:19:15: It has human adaptability.

00:19:16: Are you just upgrading?

00:19:17: Your software or are you upgrading the operating model of your people?

00:19:21: If you enjoyed this episode, new episodes drop every two weeks.

00:19:24: Also check out our other editions on cloud defense tech digital products and services artificial intelligence sustainability in green ICT Defense Tech and HealthTech.