Dylan Seychell
1.5K posts
Dylan Seychell
@dylanseychell
Lecturer in AI @UMmalta. Computer vision, Applied ML. I build AI systems and advise on what works. Leading Malta's AI Literacy Programme's academic direction.
Malta Beigetreten Mayıs 2009
322 Folgt630 Follower
This is significant.
Qwen@Alibaba_Qwen
🚀 Introducing the Qwen 3.5 Small Model Series Qwen3.5-0.8B · Qwen3.5-2B · Qwen3.5-4B · Qwen3.5-9B ✨ More intelligence, less compute. These small models are built on the same Qwen3.5 foundation — native multimodal, improved architecture, scaled RL: • 0.8B / 2B → tiny, fast, great for edge device • 4B → a surprisingly strong multimodal base for lightweight agents • 9B → compact, but already closing the gap with much larger models And yes — we’re also releasing the Base models as well. We hope this better supports research, experimentation, and real-world industrial innovation. Hugging Face: huggingface.co/collections/Qw… ModelScope: modelscope.cn/collections/Qw…
English
Yes. The paradox multiplies with students becoming pros because they're being denied internships or entry-level jobs because of AI.
God of Prompt@godofprompt
weird pattern nobody's talking about the fastest ai learners right now aren't developers or students they're the people between jobs with 14 hours a day to experiment employment has become the biggest bottleneck to staying employable
English
New systematic review: UAV-based litter detection 🚁🚮
✅ 9 public datasets analysed
✅ 18 model architectures compared
✅ Georeferencing, tiling, augmentation strategies reviewed
✅ We coin the term 'Litter Management Systems'
If you're working on aerial waste detection, this is your starting point (currently for free)
📄 rdcu.be/e5NRk
#ComputerVision #DeepLearning #UAV
@UMmalta @mt_xjenzamalta @MatthiasBar6
English
Dylan Seychell retweetet
Bookmark this. This is what future proof organisations are doing. Let's make the best out of it.
PS: the same is true for education. Educators are now officially facilitators of learning.
Logan Kilpatrick@OfficialLoganK
officially giving up my “PM” title, we are all members of the technical staff now, time to embrace it
English
@rickjordandev @BrettFromDJ It can feel like cheating, but when done well, it's pure efficiency. uxplanet.org/why-does-using…
English
@BrettFromDJ built a tiny script with claude that auto-summarizes my daily slack threads into a clean email digest—saves me 45 minutes every morning and feels like cheating
English
The magic of Claude isn't in building the next billion-dollar SAAS.
That's where most people get it wrong.
It's in building private, personal products that only you use that speed up your workflow and add quality to your life.
And for that, it's absolutely magical.
English
@signulll It is all about better decision making. I recently discussed this in detail here: uxplanet.org/why-does-using…
English
was talking to someone at a bar in nyc recently.
she was an author, so inevitably given what i do (& against my will) we ended up talking about ai. when i tell you that her idea of it was so dystopian, my god. like she was visibly repulsed by it. not even skeptical. just pure visceral hate. her position was that it should be banned by the government & no one should use it.
this might be half or more of america.
Kevin Roose@kevinroose
i follow AI adoption pretty closely, and i have never seen such a yawning inside/outside gap. people in SF are putting multi-agent claudeswarms in charge of their lives, consulting chatbots before every decision, wireheading to a degree only sci-fi writers dared to imagine. people elsewhere are still trying to get approval to use Copilot in Teams, if they're using AI at all. it's possible the early adopter bubble i'm in has always been this intense, but there seems to be a cultural takeoff happening in addition to the technical one. not ideal!
English
@cryptoteambuild It's not cheating. It's better decision making. uxplanet.org/why-does-using…
English
5 Claude Code agent use cases that feel like cheating—in a good way. Here's what I learned: #ClaudeAI #CodeAgent
amplifyaiworkshop.blog/claude-code-ag…
English
@Yuchenj_UW I believe it's really common. I reflected upon it and expanded my thoughts here: uxplanet.org/why-does-using…
English
“It feels like I’m cheating.”
Don’t feel that way, Andrej.
We never loved typing every line of code. We loved building cool things.
English
@saastrash Yes, it can feel that way, but I think it's deeper than that. I shared a few thoughts here: uxplanet.org/why-does-using…
English
AI is helping us build our enterprise vertical accounting application SaaS, and it's been the best PM & Engineer sidekick.
feels like cheating tbh
English
English
@BagCalls @PortaraOfficial ai trading feels like cheating in the best way
English
Trading perps manually feels like offering yourself up as exit liquidity for the algorithms
I’ve been testing @PortaraOfficial lately, and the AI integration is a total level-up for the terminal experience.
It’s not just a UI; it’s an intelligent layer that actually helps you execute better.
The future of perps isn't just decentralized...it's autonomous.
Keep an eye on this one. Apply for the beta here: portara.app/?ref=bagcalls (t.co/PWLQawpNKl)
le.hl@0xleegenz
Congratulations to my partner @PortaraOfficial for the first 5,000 users have already registered on their beta Traditional perp DEXs are good but not good enough. That’s why Portara is here to fix what the missed > AI Copilot Insights: Trade with precision > Automated Strategies help you optimize your execution > Elite Engineering: Built by industry leading dev > Cross chain swaps, no more bridging It’s time to onboard to Portara, you won’t regret it Register here: portara.app/?ref=lee
English
The "AI cheating" guilt misses the point.
Like Google Maps didn't kill navigation. It freed us for better destinations.
AI decouples prediction (what it excels at) from judgment (our human edge) and thus opens huge opportunities to rethink decision-making in classrooms, offices, and beyond.
As an educator, this is what excites me most.
Read: uxplanet.org/why-does-using…
English
After two years of AI hype, 2025 brought the correction we needed.
Markets sobered up, businesses questioned ROI, and the industry is consolidating.
I've written about what this means for 2026, the year we shift from AI magic to metrics.
@dylanseychell/sobering-up-about-ai-and-the-shift-from-magic-to-metrics-93d056dbcfe9" target="_blank" rel="nofollow noopener">medium.com/@dylanseychell…
English
And this is exactly why as educators we either change the approach in class or have students deconstruct what was everyone's comfort zone.
Education needs more perimeters and less parameters.
Let's truly give the classroom to our students.
François Fleuret@francoisfleuret
TFW your kids ask an LLMs to summarize their lecture notes and the model adds comments about how the lecturer confuses certain points and is overall unclear and messy.
English
@shiri_shh Cool video. This is a more detailed account of how we got here: uxplanet.org/google-is-back…
English
"95% of generative AI pilots deliver zero measurable ROI", @MIT Project NANDA 2025 report.
"$644B spent globally—mostly infrastructure", @Gartner_inc 2025.
Two lessons from 2025 to remind us to hype less and build more effectively.
UX and/for/of AI
English
English
This is hands-on AI and engineering with immediate environmental impact. We're building something that works in a team with a proven track record.
@UMmalta @MatthiasBar6
Funded by @mt_xjenzamalta
Full details & application in comment👇
English
In @UMmalta Project AWIGS, we detected litter from aerial imagery. Now it's time to pick it up. 🚁
Project TALC is recruiting a mechanical engineer to build the robotic gripper that will retrieve litter from difficult-to-access areas across Malta.
Apply by 12 Jan 2026 👇
English