Best of LinkedIn: ICT & Tech Insights CW 31/ 32

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

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.

00:00:19: You can find more info in the description Absolutely So.

00:00:22: imagine someone breaks into your house today

00:00:25: Okay

00:00:26: And they completely ignore you television.

00:00:28: They walk, you know right past your jewelry on the dresser and they just go straight for the heavy steel safe that's bolted to the floor.

00:00:35: Right?

00:00:35: They know they can't open it-they don't even try to crack the combination... ...they just haul it out their garage leave its sitting there

00:00:41: Just waiting.

00:00:41: Exactly Yeah Because in about say five years a master savecracker is gonna be born who could open it in seconds.

00:00:49: And uh.. That scenario isn't from some heist movie!

00:00:54: That Is The Exact Reality Of Global Cybersecurity Today.

00:00:57: It really is.

00:00:58: it's the defining threat model of our time, honestly.

00:01:01: and what makes us so urgent for you—the tech professionals listening–is that... The timeline for that master safecracker to arrive… Is well its compressing much faster than industry anticipated.

00:01:14: We are moving entirely out theoretical physics And straight into real engineering realities and enterprise budget cycles

00:01:23: Which exactly our mission deep dive into the source material.

00:01:27: Okay, let's unpack this because we are exploring a massive convergence today

00:01:32: A total collision

00:01:33: right?

00:01:33: We're looking at how quantum computing breakthroughs AI driven cyber threats and next generation infrastructure Are all just colliding to rewrite the rules right now.

00:01:42: Yeah, you got a fascinating stack of insights from industry leaders And The main takeaway is that you can no longer afford to view these technologies as like Separate silos.

00:01:53: No, not at all.

00:01:53: They're a single continuous ecosystem

00:01:56: and if we want to understand that Ecosystem We really have to start with the catalyst right?

00:02:07: Yeah I'm genuinely floored by how fast this is moving out of the laboratory.

00:02:11: yeah like there was an incredible highlight shared by Christian B regarding a German startup called Saxon Q.

00:02:19: Oh, yeah!

00:02:20: That one is

00:02:21: huge... Right they just crossed the physical barrier that has you know bottlenecked this industry for decades.

00:02:28: They achieved over ten cubits with a ninety-nine point nine eight percent single qubit fidelity

00:02:34: Which was great on its own.

00:02:36: Yeah

00:02:36: but the absolute kicker here Is what they did at room temperature.

00:02:39: right and What's fascinating Here?

00:02:41: This changes The entire economic model of quantum computing.

00:02:44: How so

00:02:45: Well, up until very recently the foundational assumption was that functional quantum hardware required um mill kelvin temperatures.

00:02:52: Right right

00:02:53: you essentially had to build a cryogenic freezer That kept the quantum ship colder than the vacuum of deep space Whoa just to stop thermal noise from destroying The fragile quantum state of the quibits.

00:03:04: all right i've seen pictures Of those massive multi-million dollar rigs.

00:03:07: it looked like giant gold chandeliers

00:03:09: exactly the chandelier's.

00:03:10: and that Cooling requirement was the single biggest cost barrier.

00:03:15: It meant quantum computers could only exist in highly specialized custom-built labs, right?

00:03:21: But what Sex on Q has done is utilize lab grown diamond defects.

00:03:26: wait Diamond defects.

00:03:27: Yeah Mechanically they are using what are called nitrogen vacancy centers.

00:03:32: Okay

00:03:33: it's basically placing a nitrogen atom In a diamond crystal lattice Right where a carbon atoms should be.

00:03:40: Oh, interesting.

00:03:41: And that rigid diamond structure protects the quantum state of the trapped electron so well that it operates at room temperature?

00:03:49: So you could theoretically plug this thing straight into a standard wall

00:03:53: outweigh... Any much!

00:03:54: Yeah

00:03:54: It means quantum hardware can finally start to look and act like standard rack-mounted equipment in normal data center.

00:04:01: Precisely While startups are solving the deployment friction The heavy weights are accelerating the timeline.

00:04:08: You know, Kerenicus noted in his insights that IBM is publicly targeting demonstrated quantum advantage by twenty-twenty six.

00:04:14: Twenty-twty six?

00:04:15: That's right around the

00:04:16: corner Yeah And by twenty twenty nine.

00:04:18: they're aiming to deploy a fully error corrected system.

00:04:21: They are calling Starling Wow So we aren't talking about a distant hazy horizon anymore.

00:04:26: Twenty twenty nine Is just what one or two enterprise hardware refresh cycles away.

00:04:31: Okay I hear twenty twenty six and twenty twenty Nine and honestly my Hardware hype cycle alarm Starts ringing.

00:04:38: a little bit

00:04:38: fair enough

00:04:39: because we have been hearing about quantum being you know five years away for A decade now.

00:04:44: Oh absolutely the classic joke,

00:04:46: right?

00:04:46: So why is this any different from just cramming more physical noisy quibits onto a chip?

00:04:51: Because I mean if you can't actually control The errors a thousand noise equipments Just gives You a bigger More expensive random number generator.

00:04:59: That is the exact right skepticism to have.

00:05:01: Okay, and it's precisely why The entire focus of the industry has pivoted from just chasing raw quibbit counts To solving algorithmic efficiency.

00:05:09: And crucially error correction Makes sense.

00:05:11: It's not about hardware alone anymore.

00:05:13: Christus Four highlighted some remarkable new research From NVIDIA Where they introduced a new family Of quantum LDPC codes.

00:05:21: LDPC Yeah

00:05:22: that stands for low density parity check.

00:05:25: Ok.

00:05:25: so how does actually change math on errors?

00:05:28: like practically speaking.

00:05:30: Well, historically we use something called surface codes to oversimplify it.

00:05:35: Surface codes require you to arrange physical quibits in a massive two D grid and they can only talk to their immediate neighbors to check for errors.

00:05:44: the overhead for that is astronomical.

00:05:47: To protect just a hundred and ten logical working quibits, you might need over seventy thousand physical quibbit.

00:05:53: Seventy-thousand?

00:05:54: So it's like having a stadium of seventy thousand security guards to protect the VIP section of one hundred and Ten people?

00:05:59: Yes!

00:05:59: That is totally inefficient

00:06:00: Exactly.

00:06:01: But LDPC codes completely rewire that approach.

00:06:04: Instead of rigid grid they use highly interconnected webs which allows for long range connections between quibbits.

00:06:10: This new code structure protects those exact same hundred.

00:06:16: Are

00:06:16: you serious?

00:06:17: Yeah.

00:06:17: You dropped from seventy thousand down to under a thousand.

00:06:20: That is a staggering reduction in hardware requirements.

00:06:23: It really is.

00:06:24: And crucially, NVIDIA using classical GPUs To decode and correct these errors In just One point two to two point two milliseconds.

00:06:34: Wow, because if you cannot correct the quantum errors in real time The entanglement just collapses.

00:06:39: so we actually have classical supercomputing Directly saving quantum systems from their own noise.

00:06:45: that

00:06:45: is such a fascinating symbiosis right?

00:06:48: and We're seeing this speed and fidelity jump into other modalities too.

00:06:52: yeah like albert shared That d-wave Just published A pure reviewed paper In nature detailing a ninety nine point nine percent Fidelity two quibbit entangling gate, and it runs in about five hundred nanoseconds.

00:07:03: Which is unbelievably fast!

00:07:05: Yeah...and preserves the native ability to detect common errors right at the hardware level.

00:07:10: So the math and physics are finally aligning?

00:07:13: Fault tolerance really no longer just theoretical.

00:07:15: It's incredible but you know its easy.

00:07:17: get bogged down on a nanosecond.

00:07:18: than the parity checks Oh for

00:07:19: sure..its very dense

00:07:21: Here where gets interesting though There was an anecdote in the sources that profoundly grounded this for me.

00:07:27: Oh,

00:07:28: The Summer School Story?

00:07:29: Yes!

00:07:30: Bilal A shared this story from a IBM Qiskit summer school... ...a young child whom he affectionately called Schrodinger's Kid.

00:07:39: She wanted to run real quantum job.

00:07:41: I love it.

00:07:42: So she coded four qubits together using one Hadamard gate and three CNOT gates.

00:07:48: A very classic entanglement setup.

00:07:51: Yeah,

00:07:51: when asked to explain what she just built She didn't use any jargon.

00:07:54: Right!

00:07:55: She called them four magic switches And said if you flip the first one The rest of them will always match it every single time.

00:08:01: Just like best friends Best

00:08:02: Friends.

00:08:03: It's a beautifully simple way To describe quantum entanglements.

00:08:07: It really is...and she intuitively understood that changing this state Of One instantly defines This State of Others.

00:08:13: Yeah And then she picked a live IBM quantum computer, Ibn Fez hit submit on her laptop and watched the job run over a thousand times.

00:08:21: On

00:08:21: real hardware?

00:08:22: On Real Hardware!

00:08:23: She saw results mostly all zeros or all ones with just tiny bit of noise in middle.

00:08:28: It's incredible

00:08:30: But think about implications for that.

00:08:31: second.

00:08:32: If child can casually open a laptop write a script and run entangled quantum states on real physical hardware halfway across the world today.

00:08:44: What does that mean for highly sophisticated, state-sponsored threat actors?

00:08:48: It means we are already living inside a massive cybersecurity crisis.

00:08:52: The timeline we just established directly triggers an immediate vulnerability in our current digital infrastructure.

00:09:01: And this brings us right back to that Safe Cracker analogy you opened with, it's a threat model known as Harvest Now decrypt later which was heavily emphasized by Steve Dees and Chelsea Larson Andrews.

00:09:13: If you are managing sensitive data right now, this is the part that should definitely keep you up at night.

00:09:17: Without a doubt!

00:09:18: Thread actors whether they're nation states or advanced syndicates... They currently siphoning and storing vast oceans of encrypted data.

00:09:26: Right

00:09:26: just scooping it up

00:09:27: Exactly…They know perfectly well That they cannot break traditional RSA or ECC encryption today.

00:09:35: Right, yeah.

00:09:35: But they simply don't

00:09:36: care!

00:09:37: They are harvesting it now knowing that a quantum computer capable of running Shor's algorithm to shatter that encryption is only a few years away.

00:09:45: So if your organization has proprietary code or long-term state secrets Or biometric health data That still needs to remain secret in say.

00:09:55: It is effectively already compromised today.

00:09:57: The clock has already run out, and this isn't just industry paranoia by the way... ...the federal government recognizes the severity of this.

00:10:04: They've set strict post-quantum cryptography or PQC migration deadlines.

00:10:10: We're looking at December thirty first twenty thirty and twenty thirty one for all high impact federal systems to transition.

00:10:16: But the wild thing is, we don't even have to wait for a fault-tolerant quantum computer... ...to come online and see the catastrophic dangers of bad cryptographic hygiene.

00:10:25: No!

00:10:25: Not at

00:10:25: all!!

00:10:26: The vulnerabilities are literally already being exploited today.

00:10:28: Yes.

00:10:29: Dennis Mandage provided a really sobering real world example of this –the recent cold card hardware wallet breach–.

00:10:35: Oh man….

00:10:36: Yeah.. This resulted in total loss over eighty nine million dollars.

00:10:39: Okay but wait A hardware wallet is supposed to be the absolute gold standard for security.

00:10:45: It's entirely air-gapped.

00:10:47: How does an air gap device bleed eighty nine million dollars?

00:10:51: It failed because it relied on a flawed premise for generating security keys.

00:10:56: Okay, it defaulted to software-based pseudo random number generator.

00:11:02: Here is the mechanical reality of cryptography If your random numbers aren't truly fundamentally random Your private keys can be guessed.

00:11:10: Software algorithms are just math formulas.

00:11:13: They're deterministic

00:11:14: Meaning if you know starting conditions You calculate exact output every single time.

00:11:19: Exactly Often, these pseudo-random generators use something like the exact millisecond on a system clock as their seed to start math.

00:11:29: Attackers used AI accelerated algorithms to brute force and reverse engineer those predictable clock timers And they reconstructed private keys completely offline.

00:11:38: It proves that relying on classical math formulas rather than the pure physical unpredictability of true quantum randomness is just a fatal flaw when attackers have immense computational

00:11:50: power.

00:11:51: That really highlights something crucial, if quantum computing is the future master safecracker... ...that will eventually open The Vault?

00:11:59: Yeah.

00:12:00: Artificial intelligence is the crowbar they're using to rip The Vault out of the floor today!

00:12:07: So how is AI fundamentally altering the daily reality for security defenders right now?

00:12:12: Well, AI is completely upending the economics of a cyber attack.

00:12:17: Alexander Leslie made a profound point in his analysis.

00:12:20: he said AI makes persistence virtually free.

00:12:23: Free persistence.

00:12:24: Break that down mechanically.

00:12:26: What does it look like on the network?

00:12:27: Okay, so historically launching a cyber attack required human time patience and willingness to absorb The cost of failure right

00:12:34: sitting at a keyboard

00:12:35: exactly if A human hacker tries a digital door And its locked they have To spend their own manual Time looking for window or checking the basement.

00:12:43: It takes hours.

00:12:44: AI agents removed that human bottleneck entirely.

00:12:47: Leslie pointed to an evaluation where an autonomous AI agent carried out roughly seventeen thousand six hundred actions over a few days.

00:12:55: Seventeen thousand six-hundred?

00:12:57: Yeah, and the vast majority of those actions failed.

00:12:59: But the agent doesn't get tired.

00:13:01: No Doesn't clockout at five p.m..

00:13:03: And it doesn't cost the attacker a dime in labor.

00:13:06: Right

00:13:06: It can afford to fail seventeen thousand six hundred times for free just executing relentless low-level reconnaissance until it finds the one unpatched legacy system.

00:13:17: Or, The One Forgotten Employee Permission That Works Exactly!

00:13:20: It's like replacing a single human burglar trying to pick a lock with a swarm of microscopic mechanical termites that are trying every possible combination on every single door and window simultaneously.

00:13:31: Yes And the swarm doesn't charge an hourly rate.

00:13:33: That is a perfect analogy.

00:13:35: and because the attackers cost of trying again, it's practically zero.

00:13:38: Defenders are facing massive compression in what Leslie calls time to conviction.

00:13:44: Security

00:13:44: teams can no longer afford take three days to analyze weak signals into logs to figure out if an attack is happening

00:13:50: Because by the time human analyst connects dots

00:13:53: The AI swarm has already pivoted laterally ten times elevated privileges and encrypted network.

00:14:02: And hey, just a quick note for those of you managing IT budgets and trying to stay ahead of these exact kinds of rapid fire AI driven swarms.

00:14:10: Make sure you subscribe so you catch our future deep dives because this landscape is shifting literally weekly.

00:14:15: it really does.

00:14:16: but getting back to the defense side if Ai makes attacking essentially free And quantum computing is eventually going to break our baseline encryption.

00:14:26: How does an enterprise actually survive this?

00:14:28: Are we just inherently

00:14:29: doomed?".

00:14:29: Well, we aren't doomed.

00:14:30: but the entire cybersecurity industry has having to aggressively shift its core philosophy because absolute prevention is no longer guaranteed.

00:14:37: I mean you cannot block and AI agents seventeen thousand times a day forever right.

00:14:41: so The focus is pivoting heavily toward resilience in recovery.

00:14:45: Bill O'Connell brought up of vital metric for this new era which is meantime-to-clean recovery or MTCR.

00:14:52: Meantime to clean recovery, how does that practically differ from just you know having a really good automated backup system?

00:15:00: Because merely having backups is not a recovery strategy.

00:15:03: Fair enough!

00:15:04: Heath Renfrow emphasized this beautifully in the sources.

00:15:07: he pointed out that when you are in the middle of a catastrophic ransomware incident The question is rarely do we have back

00:15:14: ups right.

00:15:14: everyone has backups

00:15:15: exactly are much harsher.

00:15:20: Like whether the backups were actually clean?

00:15:22: Exactly, Are those back-ups fully intact or they completely air gapped and isolated from threat actor who frankly has probably been living inside your network for past six months?

00:15:34: Right because if AI agent corrupted your backup three month ago

00:15:38: Restoring them just restores the attacker.

00:15:40: No that's a nightmare

00:15:41: It is.

00:15:42: And crucially can you restore at an enterprise scale before business fundamentally collapses?

00:15:47: Time as money.

00:15:48: A backup that you can't cryptographically trust, or one that takes three weeks to restore across five thousand servers is useless.

00:15:56: Clean recovery means proving mathematically that you could bounce back into an uncompromised state rapidly as a tested business continuity function.

00:16:05: So if we pullback and look at the sheer scale of what were just discussed It's

00:16:09: massive!

00:16:10: You have agentic AI running tens of thousands attacks per day Quantum workloads demanding real-time error correction in milliseconds.

00:16:19: Yeah, you have massive enterprise wide Clean data recovery operations that need to happen in hours.

00:16:27: yeah I mean the underlying pipes of the internet simply cannot handle that.

00:16:30: no they can't.

00:16:31: The physical infrastructure has to be undergoing a radical overhaul just to support this level of compute.

00:16:36: It's a multi-billion dollar overhaul, you absolutely cannot run next generation compute on legacy architecture.

00:16:43: The power densities and bandwidth requirements are too high.

00:16:46: Yeah Priyanka Pakhar noted that the global telecom ICT infrastructure market is projected hit two hundred eighty six point four eight billion dollars by twenty

00:16:55: thirty two.

00:16:55: That monumental growth.

00:16:57: It really is driven by the rollout of five G open ran and AI-driven automation.

00:17:03: But mechanically, how has the physical hardware like the actual servers in network changing to support this demand?

00:17:10: Because you can't just keep building bigger data centers.

00:17:13: eventually You literally run out of electricity on the local grid.

00:17:16: And that power constraint Is exactly why efficiency as a new currency.

00:17:19: right now We are seeing A total architectural shift away from centralized monolithic data centers toward highly distributed, hyper-efficient ecosystems.

00:17:30: Steven Peters provided some incredible metrics that really illustrate this.

00:17:33: look at NVIDIA's new Blackwell GPUs.

00:17:37: they deliver fifty times higher throughput per megawatt.

00:17:40: Fifty times?

00:17:41: Fifty Times and thirty five times lower cost per token compared to their previous hopper architecture.

00:17:47: Fifety

00:17:47: times the throughput for the exact same amount of electrical power.

00:17:50: That isn't an iterative update No!

00:17:52: It is a generational leap in physics.

00:17:54: It has to be.

00:17:55: And if we connect this the bigger picture, what's crucial for listeners is that this level of high-performance computing isn't just reserved for hyperscalers like Amazon or Google anymore...

00:18:06: Right it's democratizing!

00:18:07: ...it's becoming standard enterprise infrastructure.

00:18:10: Companies such as Roche and Eli Lilly are actively deploying these exact high density clusters in production environments today.

00:18:18: They're using them to simulate molecular interactions with drug discovery.

00:18:21: That's amazing.

00:18:22: And what's even wilder to me is that this infrastructure is literally leaving the planet.

00:18:27: Oh, The Space Networks?

00:18:29: Yes!

00:18:30: I was completely blown away by all of her shoes break down at NTT's C- eighty nine strategy.

00:18:35: It's

00:18:35: really cool.

00:18:36: right.

00:18:36: when most us think about space networks we just think of Starlink.

00:18:39: Yeah You know a constellation of low Earth orbit satellites acting as a router To give us internet in remote areas.

00:18:45: sure but c-eighty nine isn't entirely different beast.

00:18:48: it's an integrated space to ground architecture.

00:18:51: It's treating the stratosphere and orbital space as an extension of the enterprise edge.

00:18:56: Exactly, it combines terrestrial networks with high-altitude platform stations.

00:19:00: HPS yeah right

00:19:02: which are basically giant solar powered drones serving as permanent cell towers in the stratospheric wild.

00:19:08: then layers on Leo and GEO satellites.

00:19:12: But the real shift is that they are putting actual computing power into orbit.

00:19:16: Right!

00:19:17: They aren't just bouncing raw signals off a satellite back to Earth, they're building data centers in space... ...to process the data up there

00:19:25: Which solves massive latency and bandwidth problems Because if you have terrestrial sensors or edit devices generating terabytes of data beaming all that raw data down to ground station takes massive bandwidth.

00:19:38: Right it clogs pipes

00:19:40: But if you process it in orbit, You only have to beam down the final answer.

00:19:45: It's

00:19:45: like sending someone a finished blueprint instead of shipping them an entire warehouse of lumber and nails And asking them build themselves.

00:19:52: but its vital That all of this mind-bending infrastructure, you know the orbital data centers The Blackwell GPU clusters.

00:20:00: The quantum processors it All ultimately exists to serve human needs.

00:20:05: yes It has to translate into real tangible impact.

00:20:08: technology for technologies sake is useless.

00:20:11: it has To act as an enabler

00:20:13: and there was a brilliant example Of that Human Impact from Demetri Stefano

00:20:17: Oh?

00:20:17: The Lebanon update Yes

00:20:19: He provided an update on Lebanon, which just secured a hundred and fifty million dollars in World Bank financing for a sweeping digital transformation initiative.

00:20:28: They are building a national one-stop portal called Dalhadi Lab For Citizen Services

00:20:33: And the approach they're taking to build it is what makes it so relevant To our infrastructure discussion today.

00:20:38: Right What's amazing Is that their using low code platform Yeah Instead of hiring massive expensive teams Of developers To manually code Every single government service from scratch over the course of five years,

00:20:51: which is how it usually goes

00:20:53: right.

00:20:53: they can map out The logic of a Service use this low-code architecture and cut the deployment time from months down to just weeks.

00:21:02: And that is the perfect culmination of everything we've talked about today?

00:21:04: Yeah!

00:21:05: Low-code platforms only work seamlessly because the underlying infrastructure has become so powerful and so abstracted Exactly Because we have orbital computing, massive GPU clusters and high speed distributed networks.

00:21:19: The front end user in this case a national government doesn't have to worry about the plumbing.

00:21:24: They just turn the tap

00:21:25: Right, they can just focus on deploying critical services to their citizens almost overnight.

00:21:30: We spend so much time geeking out over the physics of quantum entanglement or the power density of GPUs...

00:21:36: we do!

00:21:37: But at the end of day when a smartly implemented low-code platform Can redefine an entire nation's ability To serve its people that is The true promise Of digital transformation

00:21:48: Precisely.

00:21:48: and That really brings us to the core takeaway.

00:21:51: for you the listener today if You are a tech professional, an IT leader or developer you have to stop viewing these fields as distinct Isolated disciplines

00:22:03: on a percent.

00:22:04: You cannot look at quantum computing AI security and physical infrastructure as separate silos.

00:22:10: They are one deeply intertwined ecosystem, right?

00:22:13: The physical infrastructure provides the compute.

00:22:16: yeah, the compute enables the AI swarms.

00:22:18: uh-huh the AI optimizes the attacks.

00:22:21: the Quantum breakthroughs shatter the encryption And the infrastructure has to be resilient enough to recover the data.

00:22:28: it is a single continuous loop of risk.

00:22:31: in innovation It is.

00:22:32: And if we look just over the horizon, this raises a highly provocative question that we really need to start preparing for.

00:22:37: Okay lay it on us.

00:22:38: We've talked about AI agents aggressively mapping and attacking networks today right?

00:22:42: I mean talk about quantum computers breaking encryption tomorrow.

00:22:45: Yeah But what happens when an autonomous AI agent is given native real-time access To a fault tolerant quantum computing network?

00:22:54: What happens when it can use quantum algorithms to instantly optimize its own attack paths and bypass cryptographic defenses in milliseconds?

00:23:02: That is a reality we need to start architecting defenses for right now.

00:23:22: Also check out our other editions on Cloud Insights and Sovereignty, Digital Products & Services, AI and Agenetic Systems, Green ICT and Sustainable AI, DefenseTech and HealthTech.