Thomas Heal (@ThomasHeal) header
Thomas Heal (@ThomasHeal) profile picture

Building an Educational Platform and Device. No problems, no fun.

Perth, Western Australia
Joined February 2026
Photo by Thomas Heal (@ThomasHeal)Photo by Thomas Heal (@ThomasHeal)Photo by Thomas Heal (@ThomasHeal)Photo by Thomas Heal (@ThomasHeal)Photo by Thomas Heal (@ThomasHeal)Photo by Thomas Heal (@ThomasHeal)Photo by Thomas Heal (@ThomasHeal)Photo by Thomas Heal (@ThomasHeal)Photo by Thomas Heal (@ThomasHeal)Photo by Thomas Heal (@ThomasHeal)
Replying to @rough__sea
It’s software engineering not software typing
30
225
29
3,450
296,582
Nothing worse than entering your work password into your personal computer
6
Thomas Heal retweeted
Congrats on your failure. Most people dont even try.
231
1,220
88
10,157
212,479
A few years ago, I was aware that my digital footprint would influence the outcomes in my life. It’s a beautiful thing how fast change can come
13
Thomas Heal retweeted
We built Australia’s own AI model, and today it’s open to the world. Agentaus 1 (@AgentausTD): 220B parameters. 84.4% on MMLU-Pro. Ranked 43rd in the world. It runs on Australian servers. Nothing ever leaves Australia, and nothing you type is used to train AI. Delete a chat and it’s gone for good. Most of the world’s AI answers to somebody else. This one answers only to you. Free to start: agentaus.com.au
398
215
221
2,690
2,386,600
Looking forward to testing this out
Flipper One is Live on Kickstarter !!!!!!!!!!!!!!!!!!!!!!!! kickstarter.com/projects/fli…
42
The Internet has returned to the Wild West
Hackers are now using LLMs to continuously scan software updates, spot newly fixed bugs, and build attacks before teams can apply the fix — all without human involvement. That’s according to the latest report from the Google Threat Intelligence Group. We're helping defenders counter automated threats with Google AI Threat Defense, which combines Gemini and other frontier models to find and fix flaws before attackers can exploit them.
1
48
Thomas Heal retweeted
The college admissions essay is in a very weird place. It's impossible to say just how many of them are now written by AI, but at this point, experts agree that it's a lot of them. And while colleges generally frown upon that practice, they're using AI to evaluate the essays, too. cnn.com/2026/11/15/tech/ai-s…
289
701
598
7,172
3,442,869
This aged beautifully
companies saving $100k a year by letting AI write their codebase are about to spend $500k hiring consultants to figure out why their entire architecture collapsed in 2027.
1
121
Thomas Heal retweeted
nobody in silicon valley in 2026 has this much aura
8
21
762
12,193
Electronics is wizard level programming/engineering
21
Fault tolerant servers
What's one Al habit beginners should build early on?
19
How many days/hours of the Microsoft outage until we panic?
1
277
Progress is better than stagnation, always
1
65
Extremely interesting research
We looked at 1.1M answers from 2,700 kids. The biggest surprise: the worked example is doing most of the heavy lifting. Strong initial instruction matters the most for the 1/3 who need it.
2
40
My first LiPo charging board comes with a JST fault. It’s a good thing Fable 5.2 has just been released
1
28
I suppose it’s time to enter manufacturing
1
165
Thomas Heal retweeted
Replying to @ThomasHeal
my entire personality is a system prompt
1
1
38
Altering an LLM’s communication is still done via the system prompt. That means there’s a ton of opportunity in truely communicative AI able to adapt to user preferences on the fly. I guess this is your reminder that we are still early
1
3
553
Has any checked in on the next generation?
2
135
There’s so much potential in LLM empowered recommender systems.
1
118