Hey {{ first_name | human }},
We know attention is important for learning, but it's just as important for LLMs too. Read on to find out why.
TL;DR: The 60 Second briefing
⚡️AI bills bite: Many companies are reporting that they are limiting the use of AI by employees as the costs are quickly rising and in some cases are proving to be more expensive than human hires.
🧪AI in Education: Microsoft’s AI in Education reports that 92% of students and education leaders and 88% of educators have used AI for school-related purposes.
🚨Ford Rehires: Ford rehires 300 engineers after the AI replacement fails to be as effective in identifying quality checks.
📚 AI+education news
🧪AI in Education > What it is: AI use in education is now mainstream: Microsoft reports that 92% of students and leaders and 88% of educators have used AI for school-related work, with use increasing over the past year.
The main gap is training, not access. Many students and educators see AI skills as important, but 77% of students and 53% of educators say they have had no formal AI training.
Academic integrity remains a major concern, with around 4 in 10 students and educators worried about it. Schools therefore need clear guidance on when AI use is acceptable, limited or inappropriate.
Why this matters: The interesting point is that Microsoft is framing the problem less as “Should schools use AI?” and more as “How do schools implement it responsibly?” That means the practical challenge for schools is shifting towards policy, staff training, curriculum integration, safeguarding, data protection and assessment design
🌍 Wider AI updates
⚡️ AI Bills Bite > What it is: AI costs are becoming a practical issue. Reuters reports that businesses are moving away from always using the most powerful model and instead asking: which model is good enough for this task?
Why it matters: As AI use grows, leaders will need to think less about “which AI tool is best?” and more about matching the tool to the job: cheap or free models for simple drafting, stronger models for complex planning, analysis or feedback.
🚨Ford rehires > What it is: Ford initially implemented AI-driven checks for quality control, but found that the systems lacked the expertise and experience of human engineers. To address this, Ford rehired over 300 veteran quality inspectors to train and mentor its AI systems.
Why it matters: Rumours of the AI job apocalypse may be overstated.
🎯AI concepts every teacher should know: 4. Attention
Last week we covered embeddings, the way that LLMs assign tokens a number.
But a word, or in this case, a number, on its own is often ambiguous.
Take bank:
The child sat by the bank and watched the ducks.
The child went to the bank to open an account.
Good readers use the surrounding words to infer meaning.
AI models need to do something similar.
An embedding tells the model what a word could mean.
Attention helps the model work out what it probably means here.
So, in:
The player banked on the opponent to make an error.
the embedding for banked gives a starting point. Attention helps the model notice player and opponent, making the one meaning more likely than others
‘Till next week.
Mr A 🦾
Help a colleague save time by sharing this newsletter; distributing these ideas helps a friend get home on time and keeps our energy focused on what matters most: great teaching.
Safety & Privacy Notice
The tools and workflows mentioned are intended for professional productivity and educational enhancement. Users must ensure that any AI implementation remains compliant with their local data protection regulations and institutional safeguarding policies.
Data Privacy: Do not enter personally identifiable information (PII), sensitive student records, or confidential institutional data into public AI models.
Verification Required: AI-generated content can be inaccurate, biased, or out of date. Always maintain a "human-in-the-loop" approach by reviewing and fact-checking all outputs before use.
Professional Judgement: These suggestions do not substitute for formal legal, clinical, or safeguarding advice. Final responsibility for accuracy and appropriateness remains with the professional user.
