The Japan Times - Inner workings of AI an enigma - even to its creators

EUR -
AED 4.035497
AFN 70.879608
ALL 90.379658
AMD 396.143424
ANG 1.967368
AOA 1008.736213
ARS 1705.722423
AUD 1.631453
AWG 1.977914
AZN 1.846469
BAM 1.955437
BBD 2.213208
BDT 137.250764
BGN 1.948762
BHD 0.414479
BIF 3302.017015
BMD 1.098841
BND 1.41625
BOB 12.818386
BRL 6.34998
BSD 1.098906
BTN 102.65987
BWP 15.382283
BYN 3.297818
BYR 21537.282362
BZD 2.21021
CAD 1.571821
CDF 2520.741106
CHF 0.924713
CLF 0.027761
CLP 1092.599755
CNY 7.325038
CNH 7.326923
COP 3636.086637
CRC 489.513548
CUC 1.098841
CUP 26.373142
CVE 110.378416
CZK 24.374505
DJF 195.285735
DKK 7.474398
DOP 66.425092
DZD 150.659607
EGP 59.025558
ERN 16.482614
ETB 177.297929
FJD 2.525026
FKP 0.833325
GBP 0.832388
GEL 2.845536
GGP 0.833325
GHS 12.90055
GIP 0.833325
GMD 81.314071
GNF 9735.730589
GTQ 8.398517
GYD 229.816733
HKD 8.621808
HNL 29.756236
HRK 7.507243
HTG 143.974573
HUF 366.533838
IDR 20063.736666
ILS 3.576337
IMP 0.833325
INR 102.714101
IQD 1664.744019
IRR 2119722.948453
ISK 136.806024
JEP 0.833325
JMD 174.62757
JOD 0.779084
JPY 175.300841
KES 142.783348
KGS 96.092572
KHR 4477.776293
KMF 492.280045
KPW 988.957228
KRW 1462.809769
KWD 0.342498
KYD 0.915838
KZT 487.903833
LAK 24765.131862
LBP 98401.206219
LKR 365.675403
LRD 185.429214
LSL 18.614474
LTL 3.244591
LVL 0.664678
LYD 7.09303
MAD 10.954071
MDL 20.121446
MGA 5043.679904
MKD 61.49832
MMK 2307.371362
MNT 3961.860404
MOP 8.881431
MRU 44.673325
MUR 52.590226
MVR 16.886464
MWK 1907.587522
MXN 20.243199
MYR 4.498216
MZN 70.214584
NAD 18.61457
NGN 1484.963102
NIO 40.404225
NOK 10.583871
NPR 164.255991
NZD 2.000209
OMR 0.422499
PAB 1.098876
PEN 3.752544
PGK 4.937917
PHP 69.959931
PKR 303.835236
PLN 4.402341
PYG 6846.72851
QAR 4.005607
RON 5.349172
RSD 117.631014
RUB 90.157894
RWF 1621.889223
SAR 4.127733
SBD 8.941314
SCR 15.183341
SDG 660.951956
SEK 11.133616
SGD 1.41617
SHP 0.831951
SLE 25.05643
SLL 23042.149052
SOS 629.084996
SRD 41.138384
STD 22743.788841
STN 24.888747
SVC 9.615258
SYP 14287.130231
SZL 18.614266
THB 37.170475
TJS 10.126934
TMT 3.845943
TND 3.343226
TOP 2.645746
TRY 55.428803
TTD 7.440686
TWD 36.179446
TZS 2911.931778
UAH 49.816449
UGX 4560.173891
USD 1.098841
UYU 44.643334
UZS 13219.056641
VES 1005.990962
VND 28253.947588
VUV 132.790424
WST 3.079727
XAF 655.957
XAG 0.023206
XAU 0.000285921949
XCD 2.969673
XCG 1.980532
XDR 0.776938
XOF 655.957
XPF 119.331742
YER 259.930263
ZAR 18.599257
ZMK 9890.885316
ZMW 22.51141
ZWL 353.826333
SSP 6534.755674
MXV 2.270971
  • RYCEF

    0.1000

    19.81

    +0.5%

  • VOD

    -0.2500

    16.33

    -1.53%

  • CMSC

    -0.2800

    20.12

    -1.39%

  • RBGPF

    -2.0000

    65

    -3.08%

  • RIO

    -0.2500

    94.16

    -0.27%

  • BTI

    -1.2000

    54.85

    -2.19%

  • NGG

    -0.3000

    74.94

    -0.4%

  • GSK

    -0.3500

    49.35

    -0.71%

  • BCE

    -0.4200

    20.14

    -2.09%

  • CMSD

    -0.3400

    19.93

    -1.71%

  • BP

    -0.9100

    43.52

    -2.09%

  • RELX

    0.1700

    33.24

    +0.51%

  • BCC

    -0.6200

    75.97

    -0.82%

  • AZN

    -1.9400

    164.21

    -1.18%

  • JRI

    0.0000

    10.77

    0%

Inner workings of AI an enigma - even to its creators
Inner workings of AI an enigma - even to its creators / Photo: Kirill KUDRYAVTSEV - AFP

Inner workings of AI an enigma - even to its creators

Even the greatest human minds building generative artificial intelligence that is poised to change the world admit they do not comprehend how digital minds think.

Text size:

"People outside the field are often surprised and alarmed to learn that we do not understand how our own AI creations work," Anthropic co-founder Dario Amodei wrote in an essay posted online in April.

"This lack of understanding is essentially unprecedented in the history of technology."

Unlike traditional software programs that follow pre-ordained paths of logic dictated by programmers, generative AI (gen AI) models are trained to find their own way to success once prompted.

In a recent podcast Chris Olah, who was part of ChatGPT-maker OpenAI before joining Anthropic, described gen AI as "scaffolding" on which circuits grow.

Olah is considered an authority in so-called mechanistic interpretability, a method of reverse engineering AI models to figure out how they work.

This science, born about a decade ago, seeks to determine exactly how AI gets from a query to an answer.

"Grasping the entirety of a large language model is an incredibly ambitious task," said Neel Nanda, a senior research scientist at the Google DeepMind AI lab.

It was "somewhat analogous to trying to fully understand the human brain," Nanda added to AFP, noting neuroscientists have yet to succeed on that front.

Delving into digital minds to understand their inner workings has gone from a little-known field just a few years ago to being a hot area of academic study.

"Students are very much attracted to it because they perceive the impact that it can have," said Boston University computer science professor Mark Crovella.

The area of study is also gaining traction due to its potential to make gen AI even more powerful, and because peering into digital brains can be intellectually exciting, the professor added.

- Keeping AI honest -

Mechanistic interpretability involves studying not just results served up by gen AI but scrutinizing calculations performed while the technology mulls queries, according to Crovella.

"You could look into the model...observe the computations that are being performed and try to understand those," the professor explained.

Startup Goodfire uses AI software capable of representing data in the form of reasoning steps to better understand gen AI processing and correct errors.

The tool is also intended to prevent gen AI models from being used maliciously or from deciding on their own to deceive humans about what they are up to.

"It does feel like a race against time to get there before we implement extremely intelligent AI models into the world with no understanding of how they work," said Goodfire chief executive Eric Ho.

In his essay, Amodei said recent progress has made him optimistic that the key to fully deciphering AI will be found within two years.

"I agree that by 2027, we could have interpretability that reliably detects model biases and harmful intentions," said Auburn University associate professor Anh Nguyen.

According to Boston University's Crovella, researchers can already access representations of every digital neuron in AI brains.

"Unlike the human brain, we actually have the equivalent of every neuron instrumented inside these models", the academic said. "Everything that happens inside the model is fully known to us. It's a question of discovering the right way to interrogate that."

Harnessing the inner workings of gen AI minds could clear the way for its adoption in areas where tiny errors can have dramatic consequences, like national security, Amodei said.

For Nanda, better understanding what gen AI is doing could also catapult human discoveries, much like DeepMind's chess-playing AI, AlphaZero, revealed entirely new chess moves that none of the grand masters had ever thought about.

Properly understood, a gen AI model with a stamp of reliability would grab competitive advantage in the market.

Such a breakthrough by a US company would also be a win for the nation in its technology rivalry with China.

"Powerful AI will shape humanity's destiny," Amodei wrote.

"We deserve to understand our own creations before they radically transform our economy, our lives, and our future."

M.Saito--JT