Up Times
What are tech manufacturers rushing to market that businesses’ operations are coming to depend on? Who’s vetting these tech products? And if adversaries wanted to exploit the underlying weaknesses, what’s standing in their way?
Dozens, hundreds, or even thousands – Recently Bryley partner Huntress announced former CISA head Jen Easterly would be taking on an advisory position. Easterly has worked to shift security responsibility back to tech manufacturers while bringing clarity to our adversaries so we can better deal with them.
Easterly in 2023: What the United States faces is less a cyber problem than a broader technology and culture problem. The incentives for developing and selling technology have eclipsed customer safety in importance – a trend that is not unique to software and hardware industries but one that has particularly pernicious effects … [Americans] have unwittingly come to accept that it is normal for new software and devices to be indefensible by design. They accept products that are released to market with dozens, hundreds, or even thousands of defects. They accept that the cybersecurity burden falls disproportionately on consumers and small organizations, which are often least aware of the threat and least capable of protecting themselves.
Software and hardware are being rushed into the marketplace while nearly all of these products are built on vulnerable legacy foundations. Meanwhile, as Easterly describes, below, this year, threats are mounting through state-sponsored criminality that exploits these weaknesses.
It’s time we stopped naming [cybercriminal] groups in ways that mystify, glamorize, or sanitize their nefarious activities. Fancy Bear isn’t a cartoon villain – it’s Russian military intelligence. Charming Kitten isn’t a meme-worthy hacker collective – it’s Iranian state-sponsored espionage. These actors don’t deserve clever names. Calling them dirtbags would frankly be more appropriate … the truth is, we should aim for accuracy over branding. And when attribution is clear, we should say so: China, Russia, Iran, North Korea. Calling them by name isn’t inflammatory – it’s clarifying for the cybersecurity community and the public it seeks to defend.
From a pencil needing to be sharpened, to a pen needing a new ink cartridge – the technologies we adopt maintain their usefulness when we take care of them.
Guide to the Uptime Mindset
The purpose of an organization is to be useful
–Ralph Waldo Emerson1
It’s hard to argue with the simple strategy of ‘fix it when it breaks.’ And when it comes to pencils and pens – absolutely. There is no risk in using these till you can’t anymore. Throwing a pencil away and picking up a new one is meaningless.
But the stakes are entirely different when it comes to how much organizations’ operations are now built on tech tools. If a server fails or becomes inaccessible or a database gets corrupted, the impact to many people in and outside of your organization can be great. And so the strategy for maintaining the operation of these tools should match the stakes … [6 min. read] Continue Reading >
1 Emerson was describing our individual lives, but the principle of having a useful thing to do holds up when we get together in groups.
One of the beautiful centerpieces from the event at Clintons Bar and Grille
Summer retirement party
In honor of Audrey Baker and Scott Gold
This summer brought the retirement of two of Bryley’s long-time associates Scott and Audrey.
Audrey Baker, client services representative, had been with Bryley more than ten years. Audrey will be missed by many colleagues, clients and partners. She and her husband run Whites Ferry Marina in Marshfield, where Audrey will be devoting herself full-time. Audrey also will be traveling, spending time with family, attending concerts, gardening, doing home projects and managing family elder care.
Scott Gold, senior network engineer, is in his twenty-fourth year with Bryley and will remain assisting Bryley clients through near the end of 2025. Among the carousel of pictures above you can see a display in honor of Scott with a little guitar. Scott plans to return more to his bass guitar, though not to the touring of his younger days … [2 min. read] Continue Reading >
They’re working on it
ChatGPT sharing has been showing up in Google searches
Google is in the process of removing captured chats from its dataset
Thousands of ChatGPT conversations were showing up in Google Search results, including chats with personal and work information.
OpenAI has removed the option to “Make link discoverable” in its “Share” feature-set and is working with Google to de-index past conversations. Even as this issue gets resolved through Google’s de-indexing, we need to be aware of and act according to the paltry privacy protections baked into chatbots.
Even if the tools feel private, they’re often not. And as this story shows, the internet doesn’t forget, even when you try to delete [that is, deleting from ChatGPT is distinct from whether or not Google has indexed the content], Amanda Caswell of Tom’s Guide writes … [4 min. read] tomsguide.com
Seems to have been small comfort to defeated Jeopardy contestants Jennings and Rutter when they later beat an IBM Watson lookalike with questions + dodgeballs
“Just don’t call me Artificial Intelligence”
The story of humanity getting whomped at Jeopardy by IBM’s Watson
After Watson won, [current Jeopardy host Ken] Jennings was crowded by IBM engineers and executives who were eager to tell him how valuable his own data had been as they had programmed the computer. “They were like, ‘You should feel great. There’s a lot of you in Watson,’ ” Jennings said. “It did not make me feel any better,” Slate Magazine reports.
Back in 2011 many of us got our first taste of what it’s like to go up against a thinking machine in the form of a Jeopardy competition between two of the game’s champions and IBM’s Watson – a roomful of parallel processor towers (the processors run simultaneously and in different modes – like having many brains at once). And our fiction writers had so warned about the dangers of Artificial Intelligence that the IBM execs wanted to call Watson anything but that.
“I started to notice something that you would never, ever, ever see on the show: The human contestants would sort of team up against Watson,” said [Todd] Crain, who hosted [practice games at IBM] over the course of 25 months. “They were hoping that we as humanity would win …” [20 min. read] slate.com (paywall)
This is not one of Norma Kamali’s AI updates on the Farrah Fawcett swimsuit, but an incredible simulation
Designer Norma Kamali went to MIT to study AI’s professional implications
Eventually Kamali trained a large language model on her fifty-seven years of fashion design
So therefore she created a closed-loop AI [instance,] trained solely on her work … [Kamali’s] AI-generated designs [are] now featured on her website alongside her traditional collections, per MIT.
This is an interesting application of large language models in the security of a proprietary, closed system with noteworthy results … [10 min. read] hbr.org
What’s thin, orange and secure? I mean besides this tie.
Apple sets its sights on thwarting iPhone spyware attacks
Components now hand off unique passwords in order to communicate
Maybe you heard it was recently the annual iPhone event. But this hopefully successful change in mobile security was only mentioned relatively quietly.
Apple worked with ARM Holdings to introduce in new iPhones what Apple has named Memory Integrity Enforcement (MIE).
MIE is designed to combat memory corruption bugs (e.g. attackers flood memory with malicious data, causing it to spill into unauthorized areas). The most common of these vulnerabilities are exploited by spyware developers like NSO’s Pegasus.
The new tech works by assigning secret passwords to memory sections, blocking unauthorized access and making it more difficult and therefore more expensive for attackers to exploit. Independent iOS app developers will need to integrate this functionality for it to work broadly.
MIE is a good thing and it might even be a big deal. It could significantly raise the cost for attackers and even force some of them out of the market, mobile security researcher Matthias Frielingsdorf said … [6 min. read] techcrunch.com
This supercomputer is the bridge from CPUs in parallel, like IBM’s Watson, to today’s generative AI, based on GPUs running in parallel (in parallel means many processors working at the same time)
One last blast from your past
Dell-branded EMC supercomputer arrived at MIT’s Lincoln Labs in 2016
Nine years ago MIT’s Lincoln Labs was excited to announce its EMC Dell supercomputer built on Intel Xeon Phi chips.
This is on a huge scale, but a publicly available demonstration of the type of computing architecture – largely parallel CPU clusters – that powered Machine Learning (a branch or Artificial Intelligence) in products like Bryley’s Advanced Email Threat Protection. Machine Learning processing occurred well before GPU-accelerated data-centers became an industry norm.
This Lincoln Labs supercomputer was replaced in 2019 by Hewlett-Packard’s TX-GAIA – based on hundreds of Nvidia GPUs and located in Holyoke … [4 min. watch] youtube.com
Note: The section directly above is Bryley’s curated list of external stories. Bryley does not take credit for the content of these stories, nor does it endorse or imply an affiliation with the authors or publications in which they appear.
Get Up Times, useful tech news by New Englanders in Your In-Box
- Subscribe to Up Times, the monthly New England-centric technology newsletter.
- Up Times covers:
-
- Trends in New England tech
- Security tips you can implement now
- Updates on regional and national laws and compliancies
- IT-related developments
- Networking and cybersecurity challenges New England business managers are facing
- In continuous publication since 2000, Up Times arrives monthly in your email box.
Sign up for Up Times to have tech news and tips delivered monthly via email
© Copyright 2025 Bryley Systems Inc. All Rights Reserved. | Privacy Policy | Accessibility Statement