Page 41 - AI & Your Career review copy

AREERS: ecome n I eveloper
Mak s AI or ag How hum n bra ns cre te AI or ag c usi g ton of at , ma ine lea ning, an ramming.
Wh ta gies doing things we have
.
W o-
i ta em k C phrase “purple squirrels eat French
toast for breakfast” into French, or ask
co c-
L t a e d t g. gi he
w g.
world is it?
wi like ChatGPT and DALL-E represent
lots of hard work by lots of people
ng
m e f more deeply into the technical under-
pinnings of AI systems can help us
understand the human origins and in-
gly cal capabilities that AI can display.
c g log pu further studies and work in the field is
something that might be right for you.
Gather Data,
and More Data
a nts Th of data fed into an AI system depends
on the job that the system is meant
to do. For example, consider Gmail’s
“Smart Reply” function that supplies
ready-written responses to incoming
emails. It required “training” on almost
240 million sample emails to develop
-
m d s make sure that the quality, volume,
sponses to actual emails.
relevance, and format of training data
EV ll ver
se CO te reliable, accurate results when it be-
EV so
ed that the AI system could read it. All
th field of data science, who understand
REVIEW COPY
OP s o rat a
ob W g i y c c phase in developing successful AI
systems.
statistical analysis and modeling, pro-
it t oses of to Beg s d ing ar what to do with it. The process of
f
s 44 START-ENGINEERING.COM
IMAGE CREATED BY DALL-E BY OPENAI