Jack, 1999(2), 2000(3)

These are my favorites from 1999:

Brain-machine interfaces
========================
Barinaga. "Turning Thoughts Into Actions." beta1, #4.
  Communication devices that translate thoughts into commands to operate
  computers, driven by brain waves detected from outside the body or by
  electrodes implanted in the brain, are now bring tested in paralyzed
  patients. Recent research at the annual meeting of the Society for
  Neuroscience suggests that more sophisticated devices allowing people to
  operate a robotic arm by thought alone may not be far behind.
Wissenschaft, Sabe. "Artificial Neurons in Vivo." beta2, #5.
  Cyborgs -- human/machine hybrids -- have long been the dream of science
  fiction authors..and some military enthusiasts.  More than ten years ago
  some studies have been done about a heavily armored fighting suit that is
  hydraulically controlled through some direct brain interface.  The project
  apparently died because the communication rate is too low and unreliable.
  Today, ten years later we have a thought translation device -- which is
  still very slow -- and some research into direct integration of artificial
  and biological neurons.

Computational Finance
=====================
Farmer. "Physics of Finance." beta3, #2.
  Farmer reviews some of the theoretical methods behind "econo-physics" and
  computational finance.  The central object of study is the log-return
  function, which tells you how much you can expect an investment at a given
  time will return a certain amount of time later.  Already in the 1960s,
  Benoit Mandelbrot and other recognized that log-return histories are better
  described by fractal curves than by traditional statistical processes.
  Today we know from millions of analyzed transactions that the story is a
  little more complex: no low dimensional chaotic attractors appear to describe
  the data, although nonlinearities (a precondition of chaos) are clearly
  present.  Instead of simple fractals, one needs multi-fractals to describe
  the scaling properties of the data.  In good physics tradition the econo-
  physicists are not too shy to introduce new technical terms like "fat tails"
  and even a "smile".  Instead of econometric models, the simulations are
  done based on models of economic agents.  As a return of this theoretical
  investment one seems to get a better understanding of phenomena like
  "clustered volatility" and "information cascades".
Gopikrisnan, Plerou, Anaral, Meyer, Stanley. "Scaling of the distribution of
financial market indices." beta5, #11.
  The distribution of fluctuations of the S&P500 index over a time scale t
  is studied by analyzing three distinct databases. Database (i) contains
  approximately 1,200,000 records, sampled at 1-min intervals, for the 13-year
  period 1984-1996, database (ii) contains 868 daily records for the 35-year
  period 1962-1996, and database (iii) contains 852 monthly records for the
  71-year period 1926-1996.  The probability distributions of returns over a
  time scale t is computed, where t varies approximately over a factor of
  104 -- from 1 min up to more than one month.  It is found that the
  distributions for t 4 d (1560 min) are consistent with a power-law asymptotic
  behavior, characterized by an exponent 3, well outside the stable Levy
  regime 0 << 2.  Consistent results are discovered in the analysis of two
  other financial market indices.


These are my favorites from 2000:

AI
==
2000.4, #5: Life and Evolution in Computers, Santa Fe Institute Working Papers
  Can we build computers that are intelligent and alive?  This question has
  been on the minds of computer scientists since the dawn of the computer age
  and remains a most compelling line of inquiry.  Some would argue that the
  question makes sense only if we put scare quotes around "intelligent" and
  "alive," since we're talking about computers, after all, not biological
  organisms..

Brain-Machine Interfaces
========================
2000.6, #7: Synchronization Properties of Brain Waves, PNAS
  Electrical brain rhythms have traditionally been divided up into different
  frequency bands that are associated with different mental states.  Best known
  among them are alpha rhythms in the 8-12 Hz range that are associated with a
  relaxed resting state.  Beta rhythms (12-30 Hz) as associated with alert
  mental activity whereas gamma rhythms (30-70 Hz) have recently found a lot
  of attention in connection with the "binding process" during e.g. the
  perception of an object where different features are integrated into a
  coherent structure.  There have been a numbe of theoretical models to explain
  the synchronization (and de-synchronization) of the tens of thousans of
  neurons that have to act together in order to be observable as an electrical
  EEG signal.  Kopell et al study a simplified neuronal model to demonstrate
  how different mechanisms can give rise to different rhythms.

Computational Finance
=====================
2000.15, #5: The Nasdaq Crash of April 2000, arXiv
  The Nasdaq fell another 10% on Friday the 14th of April 2000, signaling the
  end of a remarkable speculative high-tech bubble starting in spring of 1997.
  The closing of the Nasdaq at 3321 corresponds to a total loss of over 35%
  since its all-time high of 5133 on the 10th of March 2000.  Similarities to
  the speculative bubble preceding the infamous crash of October 1929 are
  quite striking: The belief in what was coined a "New Economy" both in 1929
  and presently made the share-prices of companies with three digits price to
  earning ratios soar.  Furthermore, these two speculative bubbles, as well as
  others, both nicely fit into the quantitative framework proposed by the
  authors in a series of recent papers.