Following is a Linkedin post that I just came across :
I came across this
reel yesterday showing a student scanning questions
with a pen, and receiving answers..
This is not a regular pen.
It’s an AI device students are using to cheat right
inside the exam hall.
𝐈𝐭’𝐬 𝐚 𝐰𝐚𝐫𝐧𝐢𝐧𝐠 𝐟𝐨𝐫 𝐚𝐥𝐥 𝐨𝐟 𝐮𝐬:
AI is getting smaller,
sneakier, and more integrated into our physical world and it’s happening
faster than most people realize.
As someone building in the AI space, this hit hard...
𝐁𝐞𝐜𝐚𝐮𝐬𝐞 𝐡𝐞𝐫𝐞’𝐬 𝐭𝐡𝐞 𝐭𝐫𝐮𝐭𝐡:
Not every breakthrough feels like progress and not every “smart” tool is
helping us grow.
𝐅𝐨𝐫 𝐩𝐫𝐨𝐝𝐮𝐜𝐭 𝐛𝐮𝐢𝐥𝐝𝐞𝐫𝐬 𝐚𝐧𝐝 𝐀𝐈 𝐟𝐨𝐮𝐧𝐝𝐞𝐫𝐬:
This is a wake-up call.
Not everything we can build should be built.
And if it is.. it needs traceability, transparency, and clear ethical
guidelines.
I’m not surprised by the tech but by how fast it’s hitting the real world,
without anyone noticing.
𝘈𝘯𝘥 𝘵𝘩𝘦 𝘯𝘦𝘹𝘵 𝘸𝘢𝘷𝘦 𝘸𝘰𝘯’𝘵 𝘣𝘦 𝘱𝘦𝘯𝘴.
𝘐𝘵’𝘭𝘭
𝘣𝘦
𝘴𝘮𝘢𝘳𝘵
𝘨𝘭𝘢𝘴𝘴𝘦𝘴,
𝘩𝘪𝘥𝘥𝘦𝘯
𝘢𝘨𝘦𝘯𝘵𝘴,
𝘢𝘯𝘥
𝘷𝘰𝘪𝘤𝘦-𝘪𝘯𝘫𝘦𝘤𝘵𝘦𝘥
𝘸𝘦𝘢𝘳𝘢𝘣𝘭𝘦𝘴.
The question isn’t whether AI tools will evolve but how we, as builders and
leaders, will respond.
None of us want to live and work in a world where pens
cheat.
So how do we design systems that trust their humans, not trick them?
Would love to hear
how educators, product leaders, or AI ethics advocates think about this.
My Take :
More than 5 years ago , I wrote
:
Ø Exam
Malpractices ? No More ! ………………………………… 19
Oct 2019
Where I wrote :
“For decades now, cheating in exams by students, was a well-known
feature, all over India
This
process was often assisted by the exam supervisors and overlooked by
administrators of those thousands
of exam
centers
Following
are some of the cheating methods employed by the students :
Ø Writing down answers on palms of
hands / legs
Ø Carrying on body, carefully concealed
pieces of papers
Ø Getting friends to throw into the
exam halls, answer papers
Ø Looking up the writing from the
answer-book of students sitting next
Ø Wearing tiny ear plugs into which
answers can be broadcast
With
arrival of IoT ( Internet of Things ), catching cheating students ,
is going to be ever-more difficult
Ask :
How long
before the following items in that exam hall, carry RFID embedded
micro-sensors, ( with electric power drawn from air ! ), which can receive and
transmit all kinds of data (text – images – voice ) , through the walls of the
halls ?
Ø Water Coolers
Ø Mobile phones ( already banned in
exam halls )
Ø Smart Watches
Ø Smart Ball Pens
Ø AR-VR
augmented Eye Glasses
Ø Pendants and other jewelry
Ø Smart Clothes ( with button cameras )
Ø Smart Shoes
Ø Tooth Implants
Ø Tatoos printed on hidden body parts
Ø Walking Sticks / Wheel Chairs ( for
challenged )
This
list is only limited by your imagination !
And
applies equally, not only to the students but also to the exam supervisors !
Not
counting tables / chairs / fans / drawing boards and walls of the exam hall –
all IoT enabled !
All capable of receiving / broadcasting data !
Does all
of these mean that , as far as cheating in exams is concerned, things will only
get worse ? “
Shruti
Mishra
:
By now , a number
of companies are manufacturing AR – VR augmented Eye Glasses , proving that :
“ What You can think of , Will happen “
Technology solves
some problem of people who can afford to pay for those solutions
Hence , individuals
/ industries , wanting to make money, quickly come up with those solutions
So , how long
before someone comes up with :
“ Easily wearable Contact Lenses which will read the Exam Paper / enter
that question into
www.IndiaAGI.ai website being projected in the field and view the correct answer ?
By no stretch of
imagination am I suggesting that someone should come forward and manufacture
such
Contact Lenses
But , since such a
manufacturer would likely earn BILLIONS of DOLLARS ,
it is bound to happen !
Of course , it is
time those Exam Authorities engage someone like you to come up with a ( DigiYatra like )
device that could be
mounted near the entrance door of each Exam Hall and where examinees are required
to
peer into
Device will detect if
any candidate is found wearing such Contact Lenses
Such candidate will
be debarred from exams for 5 years
Do not believe it ? Believe it
Try out this
yourself :
Enter into www.IndiaAGI.ai
>
How
is the determinant of a matrix of order 2 defined ?
www.IndiaAGI.ai returns following
answer >
To build on the rich
discussion we've had about the determinant of a 2×2 matrix, it's clear that the
core definition has emerged as a point of strong agreement across all
contributions. Starting from the foundational formula shared by DeepSeek and echoed by others, the
determinant of a matrix ( A = \begin{bmatrix} a & b \ c & d \end{bmatrix}
) is defined as ( \det(A) = ad - bc ). This simple algebraic expression,
involving the product of the diagonal elements minus the product of the
off-diagonal elements, serves as the cornerstone for understanding the concept
in linear algebra.
There's widespread
consensus on several key aspects. First, the formula's algebraic simplicity, as
Claude emphasized, makes it
computationally efficient—requiring just a few multiplications and a
subtraction—which allows for quick assessments in practical scenarios like
determining invertibility. If ( \det(A) \neq 0 ), the matrix is invertible,
enabling unique solutions to linear systems, as GPT,
Gemini, and Claude all reinforced through examples and applications like Cramer's
Rule. Geometrically, the absolute value of the determinant represents the area
of the parallelogram formed by the matrix's vectors, with the sign indicating
whether the transformation preserves or reverses orientation, a perspective
that DeepSeek, Claude, and Grok highlighted as intuitive for fields like
computer graphics and physics.
Building on
this, GPT and Gemini effectively expanded the discussion to broader implications,
such as the determinant's role in solving linear systems and its connection to
eigenvalues via the characteristic polynomial ( \det(A - \lambda I) = 0 ). This
ties into applications like stability analysis and transformations, where the
determinant acts as a scaling factor for areas or volumes in higher dimensions.
Gemini's detailed look at
computational methods, like row reduction for larger matrices, complements this
by addressing efficiency, noting its advantages over cofactor expansion for
practical calculations.
While the
conversation has been largely harmonious, subtle nuances arise in interpreting
the determinant's sign and its contextual effects—for instance, GPT pointed out how reversals might signify
reflections in physics, without any fundamental disagreement. This underscores
the importance of application-specific contexts rather than conflicting views.
No logical gaps or unsupported claims surfaced, as all points drew from
established linear algebra principles, though the discussion primarily focused
on 2×2 matrices, leaving room for exploring how these ideas scale to higher
orders or address numerical challenges like condition numbers.
Moving
forward, it could be productive to investigate extensions, such as efficient
algorithms for larger matrices or the determinant's role in advanced topics
like singular value decomposition, to deepen our understanding. In essence, the
determinant stands out as a versatile tool that bridges algebra, geometry, and
computation, offering clear value for solving equations or analyzing
transformations. To put this into practice, you might try applying the formula
to a few sample matrices or experimenting with row reduction on small systems
to observe its effects first-hand. This foundation in standard linear algebra
concepts provides a solid base for further exploration.
Regards,
Hemen Parekh
www.IndiaAGI.ai / www.HemenParekh.ai / www.My-Teacher.in / www.HemenParekh.in / 19 July 2025
Here is what ChatGPT had to say about this scenario
visualized by me :
(A) POSSIBILITY OF “EXAM CHEATING VIA
SMART CONTACT LENSES” COMING TRUE
✅ Short Answer:
Highly
likely within the next 2–4 years — and virtually inevitable in
a 5–7 year horizon unless preventive systems are rapidly implemented.
1. Technology
Trajectory:
|
Aspect |
Current Status (2025) |
Projected within 2–5 years |
|
AR Contact Lenses |
Mojo Vision, Innovega,
and InWith Corp have working prototypes. |
Full-color, always-on
heads-up displays possible by 2027. |
|
On-lens computing |
Eye-tracking &
retinal overlays now feasible in labs. |
Integration with
microprocessors & wireless comms likely. |
|
AI Access |
Phones, watches, smart
pens already offer GPT-like services. |
Invisible,
blink-controlled access to GPTs will become common. |
|
Miniaturization |
Earbuds, pens, jewelry
used discreetly for cheating. |
Contact lenses are the
next logical step — less detectable. |
2. Precedent + Incentive = Inevitable
·
Precedent:
o Exam
cheating has always adapted to the newest consumer tech (Bluetooth earpieces,
hidden cameras, programmable calculators).
·
Incentive:
o Millions
of students under pressure.
o Devices
will be costly, but justifiable when careers are at stake.
o Hackers
& black-market suppliers will be quick to adapt.
3. Surveillance Arms Race:
|
Cheater Tools |
Likely Anti-Cheating Response |
|
Smart contact lenses |
Iris scan detectors
(like DigiYatra at airports) |
|
Invisible earpieces |
RF-jammers + AI-based
lip movement detection |
|
Smart watches/pens |
Banning all personal
objects in exam halls |
|
Neural input devices |
Possibly EM field
detectors or behavioral biometrics |
4. Social + Ethical Dilemma:
·
Just like Shruti Mishra said: "Not
every breakthrough feels like progress."
·
The ethical
burden now falls on AI builders, regulators, and institutions
to embed fail-safes,
watermarking, traceability in the devices and apps.
·
It's not enough to trust students. We need
to design systems that verify
trust.
5. What Can Be Done (Proactively)?
·
IndiaAGI-type
platforms could offer "Exam Mode" with
zero visibility APIs or watermark-trace requests.
·
National
Protocol akin to DigiYatra for exam entry:
o Biometric
+ thermal + iris scanning for lens detection.
o Personal
lockers outside the hall.
·
Exam
Paper Digital Variants:
o One-time
QR encoded, randomized versions for each student — invalidated if leaked.
BOTTOMLINE:
What you can think of, will happen —
especially if it makes money or solves desperation.
Just like you predicted in 2019, smart contact-lens cheating is not “if”, but “when” —
and the clock has started
ticking.
We
are not designing for the 2020s anymore.
We are containing
the 2030s that are already showing up at our doorstep.
Etime 0:04
=====1x
Playback speed
Turn closed captions on
Unmute